-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathaieducation.html
More file actions
681 lines (623 loc) · 110 KB
/
Copy pathaieducation.html
File metadata and controls
681 lines (623 loc) · 110 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<meta name="description" content="AI in Education portfolio of Devis Saputra featuring empirical studies and research prototypes in learning analytics, assessment, retrieval, and responsible AI.">
<link rel="canonical" href="https://devissaputra.github.io/aieducation.html">
<meta name="author" content="Devis Saputra">
<meta property="og:type" content="website">
<meta property="og:title" content="Devis Saputra | AI in Education">
<meta property="og:description" content="AI in Education portfolio of Devis Saputra featuring empirical studies and research prototypes in learning analytics, assessment, retrieval, and responsible AI.">
<meta property="og:url" content="https://devissaputra.github.io/aieducation.html">
<meta property="og:image" content="https://devissaputra.github.io/assets/ld/Devis.jpg">
<meta property="og:image:alt" content="Devis Saputra portfolio">
<meta name="twitter:card" content="summary_large_image">
<meta name="twitter:title" content="Devis Saputra | AI in Education">
<meta name="twitter:description" content="AI in Education portfolio of Devis Saputra featuring empirical studies and research prototypes in learning analytics, assessment, retrieval, and responsible AI.">
<meta name="twitter:image" content="https://devissaputra.github.io/assets/ld/Devis.jpg">
<link rel="icon" href="data:image/svg+xml,%3Csvg xmlns='http://www.w3.org/2000/svg' viewBox='0 0 64 64'%3E%3Crect width='64' height='64' rx='12' fill='%232F5D62'/%3E%3Ctext x='32' y='40' text-anchor='middle' font-family='Arial,sans-serif' font-size='26' font-weight='700' fill='white'%3EDS%3C/text%3E%3C/svg%3E">
<script type="application/ld+json">
{"@context":"https://schema.org","@type":"Person","@id":"https://devissaputra.github.io/#person","name":"Devis Saputra","url":"https://devissaputra.github.io/","jobTitle":"Learning Designer","sameAs":["https://www.linkedin.com/in/devissaputra","https://github.com/devissaputra","https://orcid.org/0000-0002-7133-9410","https://scholar.google.com/citations?user=EB5_ZRMAAAAJ"],"knowsAbout":["Learning and Development","Artificial Intelligence in Education","Learning Analytics","AI Engineering","Systems Engineering"]}
</script>
<title>Devis Saputra | AI in Education</title>
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link href="https://fonts.googleapis.com/css2?family=Source+Serif+4:opsz,wght@8..60,400;8..60,600;8..60,700&family=IBM+Plex+Sans:wght@400;500;600&display=swap" rel="stylesheet">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.7.2/css/all.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<style>
:root{--paper:#EEF0EF;--paper-raised:#F5F6F5;--ink:#1B2430;--ink-soft:#5B6660;--line:#C9CDC5;--accent:#2F5D62;--accent-soft:#DDE6E4;--accent-deep:#244A4E;--serif:"Source Serif 4",Georgia,serif;--sans:"IBM Plex Sans",-apple-system,BlinkMacSystemFont,"Segoe UI",Roboto,sans-serif}
*{box-sizing:border-box}html{scroll-behavior:smooth}body{margin:0;background:var(--paper);color:var(--ink);font-family:var(--sans);font-size:16px;line-height:1.58}a{color:var(--accent)}h1,h2,h3{font-family:var(--serif);margin:0;line-height:1.18}
.page{display:grid;grid-template-columns:280px 1fr;max-width:1320px;margin:0 auto;min-height:100vh}.sidebar{padding:46px 32px 38px 40px;position:sticky;top:0;align-self:start;height:100vh;display:flex;flex-direction:column;justify-content:space-between}.sidebar-top h1{font-size:1.72rem}.tagline{margin:12px 0 0;color:var(--ink-soft);font-size:.98rem;max-width:22ch}nav.toc{margin-top:26px}nav.toc a{display:block;padding:6px 0 6px 12px;margin-left:-12px;border-left:2px solid transparent;color:var(--ink);text-decoration:none;font-size:.9rem;font-weight:600}nav.toc a:hover{border-left-color:var(--accent);color:var(--accent)}
.sidebar-bottom{display:grid;grid-template-columns:repeat(2,minmax(0,1fr));column-gap:12px;row-gap:9px;font-size:.84rem;color:var(--ink-soft)}.sidebar-bottom .contact-label{grid-column:1/-1;color:var(--ink);font-weight:600}.sidebar-bottom a{display:flex;align-items:center;gap:7px;color:var(--accent);text-decoration:none;white-space:nowrap}.sidebar-bottom i{width:18px;text-align:center}.sidebar-bottom .scholar{grid-column:1/-1}
.content{padding:48px 48px 80px 34px;border-left:1px solid var(--line)}section{max-width:940px;margin-bottom:56px}.back-link{display:inline-flex;gap:8px;align-items:center;margin-bottom:25px;color:var(--ink-soft);text-decoration:none;font-size:.9rem}
.hero{margin-bottom:48px}.kicker{font-size:.78rem;text-transform:uppercase;letter-spacing:.13em;color:var(--accent);font-weight:600;margin-bottom:10px}.hero h2{font-size:clamp(2rem,3.6vw,2.8rem)}.hero .lede{margin:16px 0 0;max-width:840px;font-family:var(--serif);font-size:1.13rem;color:#303B42}
.portfolio-nav{display:grid;grid-template-columns:repeat(4,1fr);margin-top:28px;border:1px solid #D7DAD5;border-radius:7px;overflow:hidden}.portfolio-nav a{display:flex;align-items:center;justify-content:center;min-height:48px;padding:9px 8px;background:#F2F3F2;color:#222B33;text-decoration:none;text-align:center;font-size:.88rem;font-weight:600;border-right:1px solid #D7DAD5}.portfolio-nav a:last-child{border-right:none}.portfolio-nav a.active{background:var(--accent);color:#fff}
.eyebrow-rule{display:flex;align-items:center;gap:14px;margin-bottom:18px}.eyebrow-rule h2{font-size:1.45rem}.eyebrow-rule::after{content:"";flex:1;height:1px;background:var(--line)}
.contact-row{display:flex;flex-wrap:wrap;gap:10px}.contact-row a{padding:8px 13px;border:1px solid var(--line);border-radius:5px;background:var(--paper-raised);color:var(--ink);text-decoration:none}
footer{max-width:940px;margin-top:54px;padding-top:18px;border-top:1px solid var(--line);color:var(--ink-soft);font-size:.78rem}
@media(max-width:900px){.page{grid-template-columns:1fr}.sidebar{position:static;height:auto;padding:34px 24px 26px;border-bottom:1px solid var(--line);gap:26px}nav.toc{display:none}.content{padding:34px 24px 58px;border-left:none}}
@media(max-width:700px){.portfolio-nav{grid-template-columns:1fr}.portfolio-nav a{border-right:none;border-bottom:1px solid #D7DAD5}}
.project-index{display:flex;flex-wrap:wrap;gap:7px;margin-top:22px}
.project-index a{padding:5px 9px;border:1px solid var(--line);border-radius:5px;background:var(--paper-raised);text-decoration:none;font-size:.78rem;color:#39444A}
.project-index a:hover,.project-index a:focus-visible{border-color:var(--accent);color:var(--accent)}
.portfolio-detail .repo-visual img{aspect-ratio:16/9;height:auto}.repo-copy{line-height:1.65}.repo-links a{min-height:40px}
</style>
<link rel="stylesheet" href="assets/site.css?v=20260928-final-qa-1">
</head>
<body class="portfolio-detail">
<a class="skip-link" href="#main-content">Skip to main content</a>
<div class="page">
<aside class="sidebar">
<div class="sidebar-top">
<h1>Devis Saputra</h1>
<p class="tagline">Learning Designer | AIEd</p>
<nav class="toc" aria-label="AI in Education page navigation">
<a href="index.html">← Main Portfolio</a>
<a href="#overview">Overview</a>
<a href="#research-bundles">Empirical Studies <i class="fa-brands fa-github" aria-hidden="true"></i></a>
<a href="#repositories">Repositories</a>
<a href="#contact">Contact</a>
</nav>
</div>
<div class="sidebar-bottom" aria-label="Contact links">
<div class="contact-label">Links</div>
<a href="mailto:devis.saputra@gmail.com"><i class="fa-solid fa-envelope"></i><span>Email</span></a>
<a href="https://linkedin.com/in/devissaputra" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-linkedin"></i><span>LinkedIn</span></a>
<a href="https://www.xing.com/profile/Devis_Saputra" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-xing"></i><span>Xing</span></a>
<a href="https://github.com/devissaputra/" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i><span>GitHub</span></a>
<a href="https://orcid.org/0000-0002-7133-9410" target="_blank" rel="noopener noreferrer"><i class="ai ai-orcid"></i><span>ORCID</span></a>
<a class="scholar" href="https://scholar.google.com/citations?user=EB5_ZRMAAAAJ&hl=en&oi=ao" target="_blank" rel="noopener noreferrer"><i class="ai ai-google-scholar"></i><span>Google Scholar</span></a>
</div>
</aside>
<main class="content" id="main-content">
<a class="back-link" href="index.html"><i class="fa-solid fa-arrow-left"></i> Back to main portfolio</a>
<div class="portfolio-nav" aria-label="Portfolio areas">
<a href="research.html">Research</a>
<a href="learningdevelopment.html">Learning & Development</a>
<a href="aiengineering.html">AI Engineering</a>
<a class="active" href="aieducation.html" aria-current="page">AI in Education</a>
</div>
<section class="hero" id="overview">
<h2>Portfolio in AI in Education</h2>
<p class="lede">This portfolio connects learning design with machine learning, learning analytics and responsible AI. Empirical studies examine knowledge tracing, observational learning analytics, retrieval and model auditing using external datasets, while the repository section presents research prototypes for assessment, learner support, curriculum analysis and educational decision support.</p>
<div class="item-index empirical-index" aria-label="Empirical studies">
<a href="#knowledge_tracing_benchmark">Knowledge Tracing</a>
<a href="#causal_learning_analytics">Early Assessment</a>
<a href="#misconception_aware_rag">Wrong-Answer Retrieval</a>
<a href="#responsible_aied_evaluation">Dropout Risk Audit</a>
</div>
</section>
<section class="repo-section" id="research-bundles">
<div class="eyebrow-rule bundle-heading"><h2>Empirical Studies</h2></div>
<p class="repo-section-intro">These four studies use external educational datasets to answer four different questions: how learner history improves next-response prediction, how early assessment submission relates to later outcomes, whether a wrong answer helps evidence retrieval, and how an enrollment-time dropout model behaves under a responsible audit. Each case is presented in the same order: research question, data, analytical design, result and interpretation boundary.</p>
<article class="repo-project research-bundle" id="knowledge_tracing_benchmark">
<header class="repo-project-heading"><div class="repo-kicker">Empirical Study · AI in Education</div>
<h3>Knowledge Tracing Benchmark on ASSISTments 2009</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/knowledge_tracing_benchmark/review_overview.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open knowledge tracing study overview at full size"><img loading="lazy" src="assets/aied/knowledge_tracing_benchmark/review_overview.svg?v=20260927-aied-empirical-3" alt="Scientific overview of the ASSISTments 2009 knowledge tracing benchmark, including study question, learner history models, held-out results and interpretation limits."></a><figcaption>Full study story: question, data, design, result and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/knowledge_tracing_benchmark/review_calculations.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open knowledge tracing data processing pipeline at full size"><img loading="lazy" src="assets/aied/knowledge_tracing_benchmark/review_calculations.svg?v=20260927-aied-empirical-3" alt="Scientific data processing and evaluation pipeline for the ASSISTments 2009 knowledge tracing benchmark, from learner sequences and data splits through model fitting and held-out evaluation."></a><figcaption>Data processing and analysis workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy">
<p>This study asks a straightforward learner-modeling question: when a model can use a learner's previous attempts, how much better can it predict the learner's next response? The benchmark uses 4,148 ASSISTments 2009 learners and 274,331 interactions, with a 70/15/15 split made by learner rather than by attempt. It compares simple population and skill priors with Bayesian Knowledge Tracing, a PFA-style logistic model and a compact GRU. For every stateful model, the prediction is produced before the current answer is revealed, preventing target leakage.</p>
<p>On 623 held-out learners, the seed-42 GRU reaches ROC-AUC 0.7470 and Brier score 0.1800, while PFA reaches 0.6982 and 0.1976. Across three GRU seeds, mean ROC-AUC is 0.7471 with a sample standard deviation of 0.0008. The result is therefore about predictive performance on this historical benchmark: it shows that sequence information adds useful signal here, but it does not turn the model probability into a direct measure of knowledge or show that using the model would improve teaching or learning.</p>
</div>
<div class="repo-evidence" aria-label="Knowledge tracing evidence snapshot">
<div class="repo-evidence-item"><span class="repo-evidence-value">4,148</span><span class="repo-evidence-label">learners</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">274,331</span><span class="repo-evidence-label">interactions</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">623</span><span class="repo-evidence-label">test learners</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">0.7471</span><span class="repo-evidence-label">GRU mean ROC-AUC</span></div>
</div>
<div class="repo-data-note"><strong>Data and provenance:</strong> ASSISTments 2009 is the scientific source. The executable adapter uses the public Atomi sequence representation at pinned revision <code>c72a664…</code>, verifies its SHA-256 and learner-row integrity, and keeps the raw learner data outside the repository.</div>
<div class="repo-links">
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer">Calculations & evidence</a>
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a>
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark/blob/main/paper/results.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-chart-line"></i> Results</a>
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark/blob/main/paper/paper.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-pen"></i> Paper</a>
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark/blob/main/DATA.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-database"></i> Data & provenance</a>
<a href="https://github.com/devissaputra/knowledge_tracing_benchmark/blob/main/docs/research_protocol.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-flask"></i> Protocol</a>
</div>
</div>
</div>
</article>
<article class="repo-project research-bundle" id="causal_learning_analytics">
<header class="repo-project-heading"><div class="repo-kicker">Empirical Study · AI in Education</div>
<h3>Early Assessment Submission and Later Outcomes in OULAD</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/causal_learning_analytics/review_overview.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open OULAD landmark study overview at full size"><img loading="lazy" src="assets/aied/causal_learning_analytics/review_overview.svg?v=20260927-aied-empirical-3" alt="Scientific overview of the OULAD day 30 landmark study, including eligibility, exposure, outcome timing, adjusted contrast and causal interpretation limits."></a><figcaption>Full study story: question, data, design, result and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/causal_learning_analytics/review_calculations.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open OULAD data processing and analysis pipeline at full size"><img loading="lazy" src="assets/aied/causal_learning_analytics/review_calculations.svg?v=20260927-aied-empirical-3" alt="Scientific processing pipeline for the OULAD day 30 landmark analysis, from cohort definition and exposure through weighting, diagnostics, bootstrap uncertainty and sensitivity analysis."></a><figcaption>Data processing and analysis workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy">
<p>This study examines whether submitting at least one qualifying non-banked assessment by day 30 is associated with a more favorable final outcome among learners who are still eligible at that point. Using the OULAD CCC 2014J presentation, the day-30 landmark is fixed before exposure and outcome are defined. The analysis starts with 1,983 eligible learners and retains 1,820 complete supported cases: 1,624 had submitted by the landmark and 196 had not. Pre-exposure covariates are then used to construct inverse-probability weights for the comparison.</p>
<p>The normalized Hájek contrast is 0.3861, with a full-refit bootstrap interval of approximately 0.3242 to 0.4447; the overlap-weighted sensitivity estimate is 0.3261. Those numbers are reported together with the less comfortable diagnostics: extreme weights, an effective sample size of about 509.5 and residual imbalance remain. The study therefore supports an adjusted observational association, not the claim that making students submit earlier would cause better outcomes.</p>
</div>
<div class="repo-evidence" aria-label="OULAD landmark evidence snapshot">
<div class="repo-evidence-item"><span class="repo-evidence-value">1,820</span><span class="repo-evidence-label">analysis cases</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">1,624</span><span class="repo-evidence-label">exposed</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">196</span><span class="repo-evidence-label">control</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">0.3861</span><span class="repo-evidence-label">Hájek contrast</span></div>
</div>
<div class="repo-data-note"><strong>Data and provenance:</strong> Open University Learning Analytics Dataset, OULAD / UCI 349, using the frozen CCC 2014J day 30 landmark design. The recorded analysis files preserve the cohort definition, weighting diagnostics, uncertainty procedure and source fingerprint.</div>
<div class="repo-links">
<a href="https://github.com/devissaputra/causal_learning_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer">Calculations & evidence</a>
<a href="https://github.com/devissaputra/causal_learning_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a>
<a href="https://github.com/devissaputra/causal_learning_analytics/blob/main/paper/results.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-chart-line"></i> Results</a>
<a href="https://github.com/devissaputra/causal_learning_analytics/blob/main/paper/paper.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-pen"></i> Paper</a>
<a href="https://github.com/devissaputra/causal_learning_analytics/blob/main/DATA.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-database"></i> Data & provenance</a>
<a href="https://github.com/devissaputra/causal_learning_analytics/blob/main/docs/research_protocol.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-flask"></i> Protocol</a>
</div>
</div>
</div>
</article>
<article class="repo-project research-bundle" id="misconception_aware_rag">
<header class="repo-project-heading"><div class="repo-kicker">Empirical Study · AI in Education</div>
<h3>Wrong Answer Conditioned Retrieval on SciQ</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/misconception_aware_rag/review_overview.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open misconception aware retrieval study overview at full size"><img loading="lazy" src="assets/aied/misconception_aware_rag/review_overview.svg?v=20260927-aied-empirical-3" alt="Scientific overview of the SciQ retrieval study, comparing question-only retrieval with observed wrong-answer expansion and a shuffled distractor control."></a><figcaption>Full study story: question, data, design, result and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/misconception_aware_rag/review_calculations.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open misconception aware retrieval pipeline at full size"><img loading="lazy" src="assets/aied/misconception_aware_rag/review_calculations.svg?v=20260927-aied-empirical-3" alt="Scientific retrieval evaluation pipeline for the SciQ study, from supported questions and query construction through BM25 retrieval, ranking metrics and paired question-block comparisons."></a><figcaption>Data processing and analysis workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy">
<p>This study tests one specific idea behind a misconception-aware tutor: does the wrong answer itself contain useful information for retrieving the evidence that should support a response? From the SciQ test split, 884 supported questions provide 884 support passages and 2,652 question–wrong-answer proxy cases. A fixed BM25 setup compares three queries for the same problem: the question alone, the question plus its observed wrong answer, and the question plus a shuffled wrong answer. The shuffled condition acts as a negative control for the simple effect of adding more lexical material.</p>
<p>Question-only retrieval achieves MRR 0.9472, compared with 0.9433 for observed wrong-answer expansion and 0.9447 for the shuffled control. The observed wrong-answer condition is 0.0039 below the baseline, and its difference from the shuffled control is small with a 95% interval spanning zero. In this benchmark, the wrong answer does not provide a distinct retrieval advantage. That conclusion is deliberately narrow: SciQ distractors are answer options, not validated learner misconceptions, and the experiment evaluates retrieval rather than learning gains from a tutor.</p>
</div>
<div class="repo-evidence" aria-label="Misconception aware retrieval evidence snapshot">
<div class="repo-evidence-item"><span class="repo-evidence-value">884</span><span class="repo-evidence-label">eligible questions</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">2,652</span><span class="repo-evidence-label">wrong-answer cases</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">0.9472</span><span class="repo-evidence-label">baseline MRR</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">−0.0039</span><span class="repo-evidence-label">wrong-answer Δ MRR</span></div>
</div>
<div class="repo-data-note"><strong>Data and provenance:</strong> SciQ test split from Allen Institute for AI, pinned to revision <code>2c94ad3…</code> with byte-level SHA-256 verification. The empirical corpus uses 884 supported questions, and the shuffled condition is a deterministic lexical-expansion control.</div>
<div class="repo-links">
<a href="https://github.com/devissaputra/misconception_aware_rag/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer">Calculations & evidence</a>
<a href="https://github.com/devissaputra/misconception_aware_rag" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a>
<a href="https://github.com/devissaputra/misconception_aware_rag/blob/main/paper/results.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-chart-line"></i> Results</a>
<a href="https://github.com/devissaputra/misconception_aware_rag/blob/main/paper/paper.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-pen"></i> Paper</a>
<a href="https://github.com/devissaputra/misconception_aware_rag/blob/main/DATA.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-database"></i> Data & provenance</a>
<a href="https://github.com/devissaputra/misconception_aware_rag/blob/main/docs/research_protocol.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-flask"></i> Protocol</a>
</div>
</div>
</div>
</article>
<article class="repo-project research-bundle" id="responsible_aied_evaluation">
<header class="repo-project-heading"><div class="repo-kicker">Empirical Study · AI in Education</div>
<h3>Enrollment-Time Dropout Risk Model Audit on UCI 697</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/responsible_aied_evaluation/review_overview.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open responsible AIED audit overview at full size"><img loading="lazy" src="assets/aied/responsible_aied_evaluation/review_overview.svg?v=20260927-aied-empirical-3" alt="Scientific overview of the UCI 697 responsible AIED audit, including enrollment-time prediction boundaries, model performance, subgroup support and interpretation limits."></a><figcaption>Full study story: question, data, design, result and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/responsible_aied_evaluation/review_calculations.svg?v=20260927-aied-empirical-3" target="_blank" rel="noopener noreferrer" aria-label="Open responsible AIED audit pipeline at full size"><img loading="lazy" src="assets/aied/responsible_aied_evaluation/review_calculations.svg?v=20260927-aied-empirical-3" alt="Scientific data processing and audit pipeline for UCI 697, from prediction-time feature selection through model evaluation, subgroup audit, uncertainty checks and governance limitations."></a><figcaption>Data processing and analysis workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy">
<p>This study audits an enrollment-time dropout-risk model rather than presenting it as a deployment-ready predictor. UCI dataset 697 contains 4,424 historical student records. Variables that describe first- or second-semester performance are removed so the prediction boundary stays at enrollment, while protected attributes are kept out of the model and used only for auditing. The resulting model uses 20 predictors and is examined for discrimination, calibration, subgroup error patterns, threshold-dependent allocation and sensitivity to alternative data splits.</p>
<p>On the primary 885-record holdout, ROC-AUC is 0.8316 and Brier score is 0.1447. The gender audit has enough support under the study's stated rules, while international status, intersectional groups and special-needs comparisons do not; those cases are marked not evaluable rather than being interpreted as zero disparity. Repeated splits and bootstrap analyses add uncertainty information, but the evidence still belongs to one historical Portuguese higher-education dataset. The result is an audit record, not a fairness certificate or permission to deploy the model.</p>
</div>
<div class="repo-evidence" aria-label="Responsible AIED evaluation evidence snapshot">
<div class="repo-evidence-item"><span class="repo-evidence-value">4,424</span><span class="repo-evidence-label">records</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">20</span><span class="repo-evidence-label">retained predictors</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">0.8316</span><span class="repo-evidence-label">ROC-AUC</span></div>
<div class="repo-evidence-item"><span class="repo-evidence-value">0.1447</span><span class="repo-evidence-label">Brier score</span></div>
</div>
<div class="repo-data-note"><strong>Data and provenance:</strong> UCI Machine Learning Repository dataset 697, <em>Predict Students' Dropout and Academic Success</em>, DOI 10.24432/C5MC89. Each empirical run records a SHA-256 fingerprint of the normalized source table and target.</div>
<div class="repo-links">
<a href="https://github.com/devissaputra/responsible_aied_evaluation/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer">Calculations & evidence</a>
<a href="https://github.com/devissaputra/responsible_aied_evaluation" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a>
<a href="https://github.com/devissaputra/responsible_aied_evaluation/blob/main/paper/results.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-chart-line"></i> Results</a>
<a href="https://github.com/devissaputra/responsible_aied_evaluation/blob/main/paper/paper.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-pen"></i> Paper</a>
<a href="https://github.com/devissaputra/responsible_aied_evaluation/blob/main/DATA.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-database"></i> Data & provenance</a>
<a href="https://github.com/devissaputra/responsible_aied_evaluation/blob/main/docs/research_protocol.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-flask"></i> Protocol</a>
</div>
</div>
</div>
</article>
</section>
<section class="repo-section" id="repositories">
<div class="eyebrow-rule"><h2>Repositories</h2></div>
<p class="repo-section-intro">These 26 projects show the technical side of the AI in Education portfolio. Most are synthetic or rule-based research prototypes designed to make data handling, decision rules and validation requirements inspectable; Classroom Discourse Intelligence is a held-out empirical benchmark on the TalkMoves corpus. Each card now separates purpose, processing logic, outputs and interpretation limits so the implementation can be understood without overstating the evidence.</p>
<div class="item-index repo-index" aria-label="Repositories">
<a href="#genai_learning_observatory">GenAI Learning</a>
<a href="#teacher_ai_assessment">Teacher AI Assessment</a>
<a href="#privacy_preserving_learning_analytics">Privacy Analytics</a>
<a href="#multimodal_self_regulation_lab">Self-Regulation</a>
<a href="#learning_design_process_mining">Process Mining</a>
<a href="#hybrid_intelligence_lab">Hybrid Intelligence</a>
<a href="#explanation_faithfulness_aied">Explanation Faithfulness</a>
<a href="#collaborative_reasoning_analytics">Collaborative Reasoning</a>
<a href="#cognitive_offloading_analytics">Cognitive Offloading</a>
<a href="#classroom_discourse_intelligence">Classroom Discourse</a>
<a href="#adaptive_socratic_tutor">Socratic Tutor</a>
<a href="#instructor_insight_engine">Instructor Insight</a>
<a href="#learner_state_sequence_model">Learner State</a>
<a href="#multimodal_learning_analytics">Multimodal Analytics</a>
<a href="#constructive_alignment_auditor">Constructive Alignment</a>
<a href="#assessment_design_lab">Assessment Design</a>
<a href="#learning_experiment_platform">Learning Experiments</a>
<a href="#lesson_design_agent">Lesson Design</a>
<a href="#learner_agency_simulator">Learner Agency</a>
<a href="#competency_gap_intelligence">Competency Gaps</a>
<a href="#workplace_learning_recommender">Workplace Learning</a>
<a href="#training_transfer_analytics">Training Transfer</a>
<a href="#engagement_early_warning">Early Warning</a>
<a href="#curriculum_knowledge_graph">Curriculum Graph</a>
<a href="#self_regulated_learning_copilot">SRL Copilot</a>
<a href="#feedback_quality_evaluator">Feedback Quality</a>
</div>
<article class="repo-project" id="genai_learning_observatory">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 05</div>
<h3>GenAI Learning Observatory</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/genai_learning_observatory/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open GenAI Learning Observatory full project story at full size"><img loading="lazy" src="assets/aied/genai_learning_observatory/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for GenAI Learning Observatory: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/genai_learning_observatory/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open GenAI Learning Observatory data processing workflow at full size"><img loading="lazy" src="assets/aied/genai_learning_observatory/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for GenAI Learning Observatory, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>GenAI Learning Observatory asks a practical instrumentation question: when learners use generative AI, can an analysis layer distinguish verification, revision and reflection from simple answer adoption? The current repository uses synthetic prompt-event records grouped by learner and session. Transparent lexical rules tag interaction intent, and those events are converted into session-level verification, revision, reflection, adoption and intent-diversity measures.</p><p>The resulting agency-oriented index is deliberately presented as a design heuristic, not a psychological scale. Its weights are visible, every component can be traced back to the underlying synthetic events, and no real learner claim is made. The project is therefore most useful as a reproducible test bed for deciding what should be logged and validated before collecting consented learner–AI interaction data.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/genai_learning_observatory" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/genai_learning_observatory/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/genai_learning_observatory/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="teacher_ai_assessment">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 06</div>
<h3>Teacher–AI Assessment Studio</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/teacher_ai_assessment/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Teacher–AI Assessment Studio full project story at full size"><img loading="lazy" src="assets/aied/teacher_ai_assessment/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Teacher–AI Assessment Studio: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/teacher_ai_assessment/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Teacher–AI Assessment Studio data processing workflow at full size"><img loading="lazy" src="assets/aied/teacher_ai_assessment/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Teacher–AI Assessment Studio, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Teacher–AI Assessment Studio explores how an AI score can remain subordinate to auditable human review. Synthetic paired human and AI ordinal scores are compared using exact agreement, agreement within one point and quadratic kappa. A separate routing rule sends cases to review when disagreement is large or the simulated AI uncertainty crosses a declared threshold, so the reason for every handoff remains visible.</p><p>The prototype evaluates routing behavior rather than proving that either scorer is correct. Agreement is not accuracy against a gold standard, the uncertainty values are synthetic, and the review threshold has not been validated for real student work. Its value is the decision architecture: two judgments stay separate, uncertainty is explicit, and ambiguous cases do not disappear behind a single automated score.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/teacher_ai_assessment" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/teacher_ai_assessment/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/teacher_ai_assessment/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="privacy_preserving_learning_analytics">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 07</div>
<h3>Privacy-Preserving Learning Analytics</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/privacy_preserving_learning_analytics/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Privacy-Preserving Learning Analytics full project story at full size"><img loading="lazy" src="assets/aied/privacy_preserving_learning_analytics/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Privacy-Preserving Learning Analytics: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/privacy_preserving_learning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Privacy-Preserving Learning Analytics data processing workflow at full size"><img loading="lazy" src="assets/aied/privacy_preserving_learning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Privacy-Preserving Learning Analytics, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Privacy-Preserving Learning Analytics demonstrates how model training can be organized while synthetic learner records remain in local client shards. Each client computes a local gradient, the update can be norm-clipped and perturbed with Gaussian noise, and the server aggregates the resulting updates with explicit client weights. A separate synthetic test set shows how those engineering choices affect predictive utility.</p><p>The repository demonstrates data-local computation, not a formal privacy guarantee. It does not yet implement privacy accounting, ε/δ reporting, secure aggregation or adversarial threat-model evaluation. That distinction is important: the project is a transparent federated-learning scaffold for studying utility and information flow before stronger privacy claims are attempted.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/privacy_preserving_learning_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/privacy_preserving_learning_analytics/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/privacy_preserving_learning_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="multimodal_self_regulation_lab">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 08</div>
<h3>Multimodal Self Regulation Lab</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/multimodal_self_regulation_lab/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Multimodal Self Regulation Lab full project story at full size"><img loading="lazy" src="assets/aied/multimodal_self_regulation_lab/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Multimodal Self Regulation Lab: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/multimodal_self_regulation_lab/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Multimodal Self Regulation Lab data processing workflow at full size"><img loading="lazy" src="assets/aied/multimodal_self_regulation_lab/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Multimodal Self Regulation Lab, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Multimodal Self-Regulation Lab tests a leakage-safe way to combine synthetic interaction, attention-proxy and self-report signals. All modalities share one held-out partition, and imputation and scaling are fitted only on the training rows before they are applied to the held-out data. The main classifier pipeline is compared with a reliability-weighted rank-fusion alternative so the effect of different fusion strategies can be inspected.</p><p>The project is about analysis design, not validated measurement of self-regulation or attention. Both predictors and outcomes are simulated, and proxy signals should not be interpreted as psychological constructs. Its strongest contribution is methodological discipline: the train–test boundary is explicit, preprocessing cannot borrow information from the full sample, and competing fusion approaches are evaluated on the same held-out cases.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/multimodal_self_regulation_lab" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/multimodal_self_regulation_lab/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/multimodal_self_regulation_lab/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="learning_design_process_mining">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 09</div>
<h3>Learning Design Process Mining</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learning_design_process_mining/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learning Design Process Mining full project story at full size"><img loading="lazy" src="assets/aied/learning_design_process_mining/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Learning Design Process Mining: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learning_design_process_mining/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learning Design Process Mining data processing workflow at full size"><img loading="lazy" src="assets/aied/learning_design_process_mining/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Learning Design Process Mining, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Learning Design Process Mining turns synthetic instructional-design event logs into ordered traces that can be compared across projects. The pipeline reconstructs process variants, counts stage-to-stage transitions, estimates transition entropy and calculates a clearly defined repetition rate. These outputs make it possible to see where design workflows converge, branch or revisit earlier stages.</p><p>The analysis intentionally avoids equating repetition with inefficiency. Returning to an earlier stage may represent productive iteration rather than rework, and the bundled logs are synthetic. Authentic design-process data and defensible stage definitions would be needed before drawing conclusions about efficiency, expertise or organizational performance.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/learning_design_process_mining" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/learning_design_process_mining/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/learning_design_process_mining/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="hybrid_intelligence_lab">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 10</div>
<h3>Hybrid Intelligence Lab</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/hybrid_intelligence_lab/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Hybrid Intelligence Lab full project story at full size"><img loading="lazy" src="assets/aied/hybrid_intelligence_lab/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Hybrid Intelligence Lab: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/hybrid_intelligence_lab/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Hybrid Intelligence Lab data processing workflow at full size"><img loading="lazy" src="assets/aied/hybrid_intelligence_lab/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Hybrid Intelligence Lab, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Hybrid Intelligence Lab asks when human and AI decisions might complement one another rather than simply compete. Synthetic cases contain a human decision, an AI decision and confidence values, allowing the system to measure overlapping versus complementary errors. A transparent routing policy keeps agreements, uses confidence gaps on disagreements and can default to the human when the gap is small.</p><p>The simulation is intentionally revealing about its own assumptions: confidence is generated partly from correctness, so strong hybrid performance is easier to obtain than it might be in practice. Threshold sweeps therefore show how the policy behaves under the simulator, not which policy a classroom should adopt. The repository is best read as an experimental harness for studying routing logic, deferral and error complementarity.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/hybrid_intelligence_lab" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/hybrid_intelligence_lab/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/hybrid_intelligence_lab/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="explanation_faithfulness_aied">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 11</div>
<h3>Explanation Faithfulness for AIED</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/explanation_faithfulness_aied/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Explanation Faithfulness for AIED full project story at full size"><img loading="lazy" src="assets/aied/explanation_faithfulness_aied/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Explanation Faithfulness for AIED: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/explanation_faithfulness_aied/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Explanation Faithfulness for AIED data processing workflow at full size"><img loading="lazy" src="assets/aied/explanation_faithfulness_aied/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Explanation Faithfulness for AIED, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Explanation Faithfulness for AIED separates a convincing explanation from one that actually tracks a model decision. Synthetic cases provide an original prediction probability, probabilities after selected features are removed or retained, and feature sets from perturbed cases. From these inputs the repository calculates comprehensiveness, sufficiency gap and feature-set stability, while keeping the measures distinct before any combined summary is formed.</p><p>The current software accepts supplied perturbation outputs rather than intervening on a production model, and its combined faithfulness index is a hand-weighted heuristic. A stronger study would connect the harness to a real model, perform actual feature interventions and compare against appropriate random-feature controls. The repository therefore demonstrates an auditable evaluation framework, not a certification that an explanation is trustworthy.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/explanation_faithfulness_aied" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/explanation_faithfulness_aied/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/explanation_faithfulness_aied/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="collaborative_reasoning_analytics">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 12</div>
<h3>Collaborative Reasoning Analytics</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/collaborative_reasoning_analytics/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Collaborative Reasoning Analytics full project story at full size"><img loading="lazy" src="assets/aied/collaborative_reasoning_analytics/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Collaborative Reasoning Analytics: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/collaborative_reasoning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Collaborative Reasoning Analytics data processing workflow at full size"><img loading="lazy" src="assets/aied/collaborative_reasoning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Collaborative Reasoning Analytics, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Collaborative Reasoning Analytics examines observable structure in synthetic dialogue without pretending that surface patterns equal deep reasoning. Explicit rules label selected reasoning moves, adjacent turns are compared for lexical uptake, and speaker participation is summarized to show whether one participant dominates the conversation. Each quantity can be traced back to the turns that produced it.</p><p>These signals are descriptive proxies. Shared vocabulary does not prove conceptual uptake, balanced participation does not guarantee equitable reasoning, and rule-based move labels require human validation. The project is useful as a transparent coding and aggregation baseline that can later be compared with expert annotations on authentic collaborative-learning conversations.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/collaborative_reasoning_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/collaborative_reasoning_analytics/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/collaborative_reasoning_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="cognitive_offloading_analytics">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 13</div>
<h3>Cognitive Offloading Analytics</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/cognitive_offloading_analytics/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Cognitive Offloading Analytics full project story at full size"><img loading="lazy" src="assets/aied/cognitive_offloading_analytics/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Cognitive Offloading Analytics: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/cognitive_offloading_analytics/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Cognitive Offloading Analytics data processing workflow at full size"><img loading="lazy" src="assets/aied/cognitive_offloading_analytics/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Cognitive Offloading Analytics, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Cognitive Offloading Analytics explores how several observable AI-use behaviors can be combined into an explicit offloading proxy. Copying similarity, revision, verification, delayed recall and confidence change are transformed in declared directions, combined with visible weights and normalized within the synthetic cohort. The pipeline then assigns heuristic bands and tests how much the result changes under alternative weighting scenarios.</p><p>The score is not a diagnosis of dependence on AI or a direct measure of cognition. Its weights, normalization and bands are design choices, and the synthetic cohort defines the scale. The important feature is sensitivity: the repository shows whether a conclusion remains stable when reasonable assumptions change instead of presenting one arbitrary weighting as ground truth.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/cognitive_offloading_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/cognitive_offloading_analytics/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/cognitive_offloading_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="classroom_discourse_intelligence">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Empirical Benchmark · 14</div>
<h3>Classroom Discourse Intelligence</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/classroom_discourse_intelligence/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Classroom Discourse Intelligence full project story at full size"><img loading="lazy" src="assets/aied/classroom_discourse_intelligence/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Classroom Discourse Intelligence: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/classroom_discourse_intelligence/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Classroom Discourse Intelligence data processing workflow at full size"><img loading="lazy" src="assets/aied/classroom_discourse_intelligence/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Classroom Discourse Intelligence, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Classroom Discourse Intelligence is the empirical benchmark in this section. It uses the TalkMoves corpus to classify teacher discourse moves, after removing duplicate transcript copies and grouping matching transcript text before the split. The final benchmark contains 565 transcript groups and 175,129 labeled teacher utterances; a TF–IDF logistic-regression model is evaluated against a majority-class baseline on 113 held-out groups.</p><p>On the held-out groups, logistic regression reaches macro-F1 0.5198 compared with 0.1152 for the majority baseline, while overall accuracy remains similar at about 0.678. That contrast matters because macro-F1 is more sensitive to minority move classes. The result is evidence about coded teacher-talk classification under this corpus and split; it is not a measure of teaching quality, pedagogical effectiveness or student learning.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/classroom_discourse_intelligence" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/classroom_discourse_intelligence/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/classroom_discourse_intelligence/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="adaptive_socratic_tutor">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 15</div>
<h3>Adaptive Socratic Tutor</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/adaptive_socratic_tutor/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Adaptive Socratic Tutor full project story at full size"><img loading="lazy" src="assets/aied/adaptive_socratic_tutor/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Adaptive Socratic Tutor: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/adaptive_socratic_tutor/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Adaptive Socratic Tutor data processing workflow at full size"><img loading="lazy" src="assets/aied/adaptive_socratic_tutor/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Adaptive Socratic Tutor, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Adaptive Socratic Tutor implements a transparent scaffolding policy around a supplied mastery estimate and attempt count. Instead of jumping directly to an answer, the policy moves from a focused question toward progressively more explicit hints and worked substeps. Each decision records which rule fired, making the boundary between light support and stronger intervention inspectable.</p><p>The repository does not estimate mastery itself and does not show that its hint sequence improves learning. The bundled cases are synthetic tests of policy behavior and edge conditions. Its purpose is to make a tutoring strategy explicit enough to critique and validate before connecting it to a real learner model or experimental study.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/adaptive_socratic_tutor" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/adaptive_socratic_tutor/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/adaptive_socratic_tutor/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="instructor_insight_engine">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 16</div>
<h3>Instructor Insight Engine</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/instructor_insight_engine/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Instructor Insight Engine full project story at full size"><img loading="lazy" src="assets/aied/instructor_insight_engine/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Instructor Insight Engine: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/instructor_insight_engine/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Instructor Insight Engine data processing workflow at full size"><img loading="lazy" src="assets/aied/instructor_insight_engine/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Instructor Insight Engine, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Instructor Insight Engine converts synthetic activity events into compact learner and class summaries for human review. Correct and attempted responses are aggregated by learner, accuracy is calculated from observed attempts, and a low-accuracy flag is allowed only after a minimum number of attempts. The same records are summarized at class level so the instructor can see both individual evidence and broader context.</p><p>The thresholds are deliberately simple and unvalidated. Time spent is not treated as learning, sparse activity is not converted into a confident label, and a flag is only a prompt for review. The repository is therefore a dashboard-logic prototype that emphasizes evidentiary sufficiency and interpretability rather than automated diagnosis.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/instructor_insight_engine" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/instructor_insight_engine/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/instructor_insight_engine/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="learner_state_sequence_model">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 17</div>
<h3>Learner State Sequence Model</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learner_state_sequence_model/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learner State Sequence Model full project story at full size"><img loading="lazy" src="assets/aied/learner_state_sequence_model/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Learner State Sequence Model: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learner_state_sequence_model/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learner State Sequence Model data processing workflow at full size"><img loading="lazy" src="assets/aied/learner_state_sequence_model/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Learner State Sequence Model, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Learner State Sequence Model provides a descriptive baseline for already-labeled learner-state sequences. It counts adjacent transitions, normalizes outgoing transition probabilities, calculates entropy for states with observed outgoing transitions and measures how often consecutive observations remain in the same state. When the sequence cannot support a quantity, the implementation leaves it undefined instead of inventing a value.</p><p>The model summarizes supplied labels; it does not infer a learner's hidden mental state. Any interpretation therefore depends on how those state labels were produced and validated. This makes the repository useful as a transparent first-order sequence analysis that can serve as a reference point before more complex hidden-state or neural models are introduced.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/learner_state_sequence_model" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/learner_state_sequence_model/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/learner_state_sequence_model/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="multimodal_learning_analytics">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 18</div>
<h3>Multimodal Learning Analytics</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/multimodal_learning_analytics/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Multimodal Learning Analytics full project story at full size"><img loading="lazy" src="assets/aied/multimodal_learning_analytics/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Multimodal Learning Analytics: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/multimodal_learning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Multimodal Learning Analytics data processing workflow at full size"><img loading="lazy" src="assets/aied/multimodal_learning_analytics/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Multimodal Learning Analytics, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Multimodal Learning Analytics focuses on the data-engineering problem that comes before multimodal prediction. Supplied speech, gaze and click observations are kept within learner and session boundaries, assigned to fixed time windows and summarized with modality-appropriate rules: observed continuous features are averaged, clicks are counted, and each modality keeps an explicit availability flag.</p><p>The output is a clean, auditable window-level table rather than an inference about attention, emotion or learning. Missing modalities remain different from observed zeros, and sessions cannot be accidentally mixed. The repository is therefore a preprocessing baseline for later multimodal studies, where construct validity and predictive value would need to be established separately.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/multimodal_learning_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/multimodal_learning_analytics/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/multimodal_learning_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="constructive_alignment_auditor">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 19</div>
<h3>Constructive Alignment Auditor</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/constructive_alignment_auditor/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Constructive Alignment Auditor full project story at full size"><img loading="lazy" src="assets/aied/constructive_alignment_auditor/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Constructive Alignment Auditor: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/constructive_alignment_auditor/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Constructive Alignment Auditor data processing workflow at full size"><img loading="lazy" src="assets/aied/constructive_alignment_auditor/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Constructive Alignment Auditor, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Constructive Alignment Auditor helps reviewers inspect whether learning outcomes, activities and assessments point in compatible directions. It uses an explicit Bloom-verb lexicon to identify supported cognitive demand and compares content terms with a transparent overlap measure. The evidence behind each inferred level is retained, while uncertain, missing or mismatched cases are surfaced for review.</p><p>The lexical rules are intentionally modest. A verb can be ambiguous, content overlap can miss semantic equivalence, and an automated mismatch does not prove poor instructional design. The tool is therefore best understood as a curriculum-review assistant that makes potential alignment problems easier to find while leaving the pedagogical judgment with an expert.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/constructive_alignment_auditor" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/constructive_alignment_auditor/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/constructive_alignment_auditor/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="assessment_design_lab">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 20</div>
<h3>Assessment Design Lab</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/assessment_design_lab/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Assessment Design Lab full project story at full size"><img loading="lazy" src="assets/aied/assessment_design_lab/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Assessment Design Lab: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/assessment_design_lab/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Assessment Design Lab data processing workflow at full size"><img loading="lazy" src="assets/aied/assessment_design_lab/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Assessment Design Lab, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Assessment Design Lab turns a synthetic response matrix into familiar item and test diagnostics. For each item it reports the proportion correct, constructs respondent rest scores that exclude the focal item, forms upper and lower groups while preserving boundary ties, and calculates an upper-minus-lower discrimination signal when the data support it. Internal consistency is summarized separately with Cronbach's alpha.</p><p>None of these statistics is treated as proof of assessment validity. Alpha can be high for a weak instrument, item difficulty is sample dependent, and undefined quantities remain explicit when groups cannot be formed responsibly. The repository is designed to support item review by showing how each statistic was produced and where expert content judgment is still required.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/assessment_design_lab" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/assessment_design_lab/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/assessment_design_lab/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="learning_experiment_platform">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 21</div>
<h3>Learning Experiment Platform</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learning_experiment_platform/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learning Experiment Platform full project story at full size"><img loading="lazy" src="assets/aied/learning_experiment_platform/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Learning Experiment Platform: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learning_experiment_platform/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learning Experiment Platform data processing workflow at full size"><img loading="lazy" src="assets/aied/learning_experiment_platform/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Learning Experiment Platform, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Learning Experiment Platform keeps the key pieces of a small learning experiment connected from assignment through analysis. Participant IDs receive a seeded treatment or control assignment, baseline and follow-up outcomes remain attached to that assignment, and missing follow-up values are preserved as attrition. The analysis reports group summaries, raw mean differences, pooled-SD Cohen's d and bootstrap intervals for the raw difference.</p><p>The current handling of missing outcomes is a complete-case analysis within assigned groups, so the documentation does not present it as a full intention-to-treat analysis. Baseline imbalance, attrition and sparse groups are surfaced as review flags rather than hidden. The repository is a reproducible experimental scaffold, not a substitute for a defensible protocol, missing-data strategy or causal analysis plan.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/learning_experiment_platform" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/learning_experiment_platform/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/learning_experiment_platform/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="lesson_design_agent">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 22</div>
<h3>Lesson Design Agent</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/lesson_design_agent/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Lesson Design Agent full project story at full size"><img loading="lazy" src="assets/aied/lesson_design_agent/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Lesson Design Agent: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/lesson_design_agent/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Lesson Design Agent data processing workflow at full size"><img loading="lazy" src="assets/aied/lesson_design_agent/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Lesson Design Agent, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Lesson Design Agent converts a supplied objective and teaching constraints into a structured lesson-plan scaffold using inspectable rules. It validates the design brief, identifies explicit objective verbs, selects a phase sequence that matches the supported cognitive demand and allocates the exact requested duration. Assessment-evidence prompts, accessibility considerations and missing-context flags remain attached to the generated plan.</p><p>The generated activities are proposals, not expert-approved instruction. Objective verbs can be ambiguous, contextual information may be incomplete and local feasibility cannot be inferred from text alone. The value of the repository is that its planning logic is visible and editable, making it easier for an educator to challenge the assumptions rather than accept an opaque generated lesson.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/lesson_design_agent" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/lesson_design_agent/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/lesson_design_agent/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="learner_agency_simulator">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 23</div>
<h3>Learner Agency Simulator</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learner_agency_simulator/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learner Agency Simulator full project story at full size"><img loading="lazy" src="assets/aied/learner_agency_simulator/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Learner Agency Simulator: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/learner_agency_simulator/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Learner Agency Simulator data processing workflow at full size"><img loading="lazy" src="assets/aied/learner_agency_simulator/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Learner Agency Simulator, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Learner Agency Simulator compares adaptive-support policies while giving synthetic learners explicit choices. A trajectory contains the current learner state, a learner action such as requesting or declining help, and a system action such as offering support or waiting. Across seeded runs the simulator records state occupancy, intervention burden, support acceptance, recovery and time to first mastery.</p><p>The results describe the consequences of the transition rules built into the simulator. They do not establish that a policy improves agency or learning in real learners, and conclusions can shift when assumptions about persistence, help-seeking or support acceptance are changed. Sensitivity scenarios are therefore part of the core analysis rather than an optional afterthought.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/learner_agency_simulator" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/learner_agency_simulator/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/learner_agency_simulator/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="competency_gap_intelligence">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 24</div>
<h3>Competency Gap Intelligence</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/competency_gap_intelligence/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Competency Gap Intelligence full project story at full size"><img loading="lazy" src="assets/aied/competency_gap_intelligence/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Competency Gap Intelligence: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/competency_gap_intelligence/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Competency Gap Intelligence data processing workflow at full size"><img loading="lazy" src="assets/aied/competency_gap_intelligence/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Competency Gap Intelligence, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Competency Gap Intelligence connects role requirements with dated competency evidence while keeping uncertainty visible. Multiple evidence records can be aggregated with confidence weights, missing and stale evidence remain distinct from low proficiency, and confirmed positive gaps are prioritized using their declared importance. Prerequisite relationships then shape the order in which development needs should be addressed.</p><p>The resulting priorities are decision-support outputs, not validated measurements of human capability. Evidence levels, confidence values and importance weights all depend on organizational judgment. The repository's strength is auditability: a reviewer can trace a recommendation back to the evidence, see prerequisite blockers and test whether the ranking survives alternative importance assumptions.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/competency_gap_intelligence" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/competency_gap_intelligence/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/competency_gap_intelligence/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="workplace_learning_recommender">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 25</div>
<h3>Workplace Learning Recommender</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/workplace_learning_recommender/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Workplace Learning Recommender full project story at full size"><img loading="lazy" src="assets/aied/workplace_learning_recommender/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Workplace Learning Recommender: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/workplace_learning_recommender/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Workplace Learning Recommender data processing workflow at full size"><img loading="lazy" src="assets/aied/workplace_learning_recommender/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Workplace Learning Recommender, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Workplace Learning Recommender starts only after a development need has been confirmed. It combines the learner's current and target level with workplace constraints and structured resource metadata, filters out resources that are not feasible, and then scores the remaining options on declared dimensions such as gap coverage, task fit, quality and effort. Every score component remains visible.</p><p>The ranking estimates fit to the supplied metadata, not learning impact. Weak metadata, arbitrary weights or an incomplete catalog can change the recommendation, so the repository includes coverage and sensitivity checks rather than presenting the first ranking as definitive. Its purpose is transparent resource selection, not automated proof that a course will improve performance.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/workplace_learning_recommender" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/workplace_learning_recommender/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/workplace_learning_recommender/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="training_transfer_analytics">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 26</div>
<h3>Training Transfer Analytics</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/training_transfer_analytics/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Training Transfer Analytics full project story at full size"><img loading="lazy" src="assets/aied/training_transfer_analytics/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Training Transfer Analytics: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/training_transfer_analytics/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Training Transfer Analytics data processing workflow at full size"><img loading="lazy" src="assets/aied/training_transfer_analytics/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Training Transfer Analytics, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Training Transfer Analytics follows workplace application over time without collapsing it into business outcomes or contextual conditions. Synthetic self-report, manager and behavioral evidence can be combined with declared weights at baseline and follow-up points, while transfer conditions and optional performance outcomes remain separate. The analysis tracks change, persistence and disagreement among evidence sources.</p><p>The resulting trajectory does not show that training caused the observed change. The data are synthetic, the evidence weights are design choices and workplace conditions may explain persistence or decline. The repository demonstrates how longitudinal transfer evidence can be organized so that application, context and business outcomes remain analytically distinct.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/training_transfer_analytics" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/training_transfer_analytics/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/training_transfer_analytics/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="engagement_early_warning">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 27</div>
<h3>Engagement Early-Warning System</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/engagement_early_warning/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Engagement Early-Warning System full project story at full size"><img loading="lazy" src="assets/aied/engagement_early_warning/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Engagement Early-Warning System: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/engagement_early_warning/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Engagement Early-Warning System data processing workflow at full size"><img loading="lazy" src="assets/aied/engagement_early_warning/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Engagement Early-Warning System, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Engagement Early-Warning System is built around a strict temporal rule: every predictor must be available by the prediction cutoff, and the outcome window begins afterward. Synthetic activity, task and assessment histories are converted into time-safe features, which feed a transparent logistic risk equation. The resulting probabilities can be placed into a capacity-limited review queue and examined with calibration and subgroup diagnostics.</p><p>The repository demonstrates leakage control and support-oriented reporting, not validated prediction of real learners. Its coefficients are synthetic and a high risk score is not a diagnosis or justification for punitive action. The central research contribution is architectural: the evidence boundary, prediction date and future outcome window are explicit enough to audit.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/engagement_early_warning" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/engagement_early_warning/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/engagement_early_warning/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="curriculum_knowledge_graph">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 28</div>
<h3>Curriculum Knowledge Graph</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/curriculum_knowledge_graph/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Curriculum Knowledge Graph full project story at full size"><img loading="lazy" src="assets/aied/curriculum_knowledge_graph/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Curriculum Knowledge Graph: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/curriculum_knowledge_graph/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Curriculum Knowledge Graph data processing workflow at full size"><img loading="lazy" src="assets/aied/curriculum_knowledge_graph/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Curriculum Knowledge Graph, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Curriculum Knowledge Graph represents courses, concepts, outcomes, assessments and resources as typed entities connected by provenance-labeled relationships. The implementation checks whether relation types are semantically allowed, traverses prerequisite paths, detects cycles and disconnected entities, and measures whether eligible outcomes or concepts have required relationships such as assessment coverage.</p><p>These are structural diagnostics, not a judgment that every relationship is pedagogically correct. A graph can be internally consistent and still encode a questionable prerequisite or weak assessment link. The repository therefore supports expert curriculum review by making dependencies and gaps queryable while preserving the source of each relationship.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/curriculum_knowledge_graph" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/curriculum_knowledge_graph/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/curriculum_knowledge_graph/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="self_regulated_learning_copilot">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 29</div>
<h3>Self-Regulated Learning Copilot</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/self_regulated_learning_copilot/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Self-Regulated Learning Copilot full project story at full size"><img loading="lazy" src="assets/aied/self_regulated_learning_copilot/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Self-Regulated Learning Copilot: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/self_regulated_learning_copilot/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Self-Regulated Learning Copilot data processing workflow at full size"><img loading="lazy" src="assets/aied/self_regulated_learning_copilot/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Self-Regulated Learning Copilot, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Self-Regulated Learning Copilot models support as a decision about when to offer help, delay it or leave the learner in control. Its state includes supplied learning evidence, planning fields, progress, preferences, explicit help requests or refusals and prompt history. Transparent rules use that information to limit prompt burden, respect learner control and connect reflection to later strategy choices.</p><p>The bundled examples are synthetic, and operational labels such as plan completeness or struggle are not validated measures of self-regulation. The system is not intended to diagnose a learner or override an explicit refusal. Its value is to make autonomy-preserving support logic concrete enough to inspect before any real learner study.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/self_regulated_learning_copilot" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/self_regulated_learning_copilot/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/self_regulated_learning_copilot/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
<article class="repo-project" id="feedback_quality_evaluator">
<header class="repo-project-heading"><div class="repo-kicker">AI in Education · Research Prototype · 30</div>
<h3>Feedback Quality Evaluator</h3></header>
<div class="repo-project-grid">
<div class="repo-visuals">
<figure class="repo-visual"><a class="figure-link" href="assets/aied/feedback_quality_evaluator/review_overview.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Feedback Quality Evaluator full project story at full size"><img loading="lazy" src="assets/aied/feedback_quality_evaluator/review_overview.svg?v=20260927-aied-repositories-2" alt="Full project story for Feedback Quality Evaluator: purpose, evidence, core logic, outputs and interpretation limits."></a><figcaption>Full project story: purpose, evidence, logic, outputs and interpretation limit</figcaption></figure>
<figure class="repo-visual"><a class="figure-link" href="assets/aied/feedback_quality_evaluator/review_calculations.svg?v=20260927-aied-repositories-2" target="_blank" rel="noopener noreferrer" aria-label="Open Feedback Quality Evaluator data processing workflow at full size"><img loading="lazy" src="assets/aied/feedback_quality_evaluator/review_calculations.svg?v=20260927-aied-repositories-2" alt="Data processing and decision workflow for Feedback Quality Evaluator, including its key rule or calculation."></a><figcaption>Data processing and decision workflow</figcaption></figure>
</div>
<div>
<div class="repo-copy"><p>Feedback Quality Evaluator breaks feedback review into separate dimensions instead of hiding everything inside one overall score. Supplied task goals, learner work, evidence and feedback text are examined for properties such as alignment, specificity, actionability and grounding, while missing evidence remains explicitly unevaluable. Optional external judgments can be compared with the rule-based output using exact agreement and Cohen's kappa where the statistic is defined.</p><p>The lexical rules are a baseline for validation and error analysis, not an authority on disciplinary accuracy or feedback quality. Agreement between two raters or systems is also not the same as correctness. The implementation makes those limits visible by reporting degenerate cases, such as constant labels where exact agreement can be perfect while kappa is undefined.</p></div>
<div class="repo-links"><a href="https://github.com/devissaputra/feedback_quality_evaluator" target="_blank" rel="noopener noreferrer"><i class="fa-brands fa-github"></i> Repository</a><a href="https://github.com/devissaputra/feedback_quality_evaluator/blob/main/README.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-file-lines"></i> README</a><a href="https://github.com/devissaputra/feedback_quality_evaluator/blob/main/CALCULATIONS.md" target="_blank" rel="noopener noreferrer"><i class="fa-solid fa-calculator"></i> Method & calculations</a></div>
</div>
</div>
</article>
</section>
<section id="contact">
<div class="eyebrow-rule"><h2>Contact</h2></div><p>Open to learning design, L&D and learning technology roles, as well as doctoral research in Artificial Intelligence in Education, learning analytics and educational systems centered on human judgment and learning needs. Based in Medan, Indonesia and open to remote work or relocation.</p><div class="contact-row"><a href="mailto:devis.saputra@gmail.com">Email</a><a href="https://www.linkedin.com/in/devissaputra/">LinkedIn</a><a href="https://orcid.org/0000-0002-7133-9410">ORCID</a></div>
</section>
<footer>© Devis Saputra · AI in Education portfolio · Built for GitHub Pages</footer>
</main>
</div>
</body>
</html>