{"paper":{"title":"An Interpretable Closed-Loop Intelligent Tutoring System for Multimodal Affective Feedback in Asynchronous Presentation Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A closed-loop multimodal ITS using traceable BARS feedback produces measurable gains in presentation skills for 204 adult learners.","cross_cats":["cs.AI"],"primary_cat":"cs.HC","authors_text":"Hung-Yue Suen, Kuo-En Hung","submitted_at":"2026-05-17T14:12:40Z","abstract_excerpt":"This paper presents an interpretable closed-loop Intelligent Tutoring System (ITS) that supports feedback-guided practice for developing on-camera oral presentation skills at scale. The system operationalizes a seven-dimensional Behaviorally Anchored Rating Scale (BARS) and implements a three-layer interpretable feedback architecture that connects rubric-aligned multimodal scoring, audience-perceived expressive diagnostics, and retrieval-augmented conversational coaching to support deliberate practice. Built on an XGBoost backbone, the ITS maps multimodal inputs (facial, vocal, textual, and oc"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"In a pre-post validation study with 204 adult learners over a 30-day practice window, participants demonstrated significant improvements across all seven BARS dimensions (Cohen's d = 0.39-0.90), with practice frequency showing a strong positive association with posttest performance after controlling for baseline scores and demographics.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The pre-post gains are caused by the ITS feedback rather than by repeated practice alone or by self-selection of motivated participants; the abstract does not describe a control group or randomization that would isolate the system's contribution.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"An XGBoost-based ITS with a three-layer interpretable feedback architecture uses multimodal features from 10,360 MOOC video segments to score seven BARS dimensions and produces measurable skill gains in a 204-learner 30-day study.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A closed-loop multimodal ITS using traceable BARS feedback produces measurable gains in presentation skills for 204 adult learners.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"49f65ffac8f75237feb171301a20cccef680002b19cd65c4a3818ff11a581f79"},"source":{"id":"2605.17468","kind":"arxiv","version":1},"verdict":{"id":"d57e6bd2-404e-4842-ab3e-39d718e95a56","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T22:29:13.207699Z","strongest_claim":"In a pre-post validation study with 204 adult learners over a 30-day practice window, participants demonstrated significant improvements across all seven BARS dimensions (Cohen's d = 0.39-0.90), with practice frequency showing a strong positive association with posttest performance after controlling for baseline scores and demographics.","one_line_summary":"An XGBoost-based ITS with a three-layer interpretable feedback architecture uses multimodal features from 10,360 MOOC video segments to score seven BARS dimensions and produces measurable skill gains in a 204-learner 30-day study.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The pre-post gains are caused by the ITS feedback rather than by repeated practice alone or by self-selection of motivated participants; the abstract does not describe a control group or randomization that would isolate the system's contribution.","pith_extraction_headline":"A closed-loop multimodal ITS using traceable BARS feedback produces measurable gains in presentation skills for 204 adult learners."},"integrity":{"clean":false,"summary":{"advisory":1,"critical":4,"by_detector":{"doi_compliance":{"total":5,"advisory":1,"critical":4,"informational":0}},"informational":0},"endpoint":"/pith/2605.17468/integrity.json","findings":[{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.1016/b978-) was visible in the surrounding text but could not be confirmed against doi.org as printed.","detector":"doi_compliance","severity":"advisory","ref_index":19,"audited_at":"2026-05-19T22:41:39.761878Z","detected_doi":"10.1016/b978-","finding_type":"recoverable_identifier","verdict_class":"incontrovertible","detected_arxiv_id":null},{"note":"Identifier '10.1109/access.2024.335678' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":42,"audited_at":"2026-05-19T22:41:39.761878Z","detected_doi":"10.1109/access.2024.335678","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null},{"note":"Identifier '10.3389/feduc.2025.145022' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":24,"audited_at":"2026-05-19T22:41:39.761878Z","detected_doi":"10.3389/feduc.2025.145022","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null},{"note":"Identifier '10.1109/tlt.2023.3323123' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":17,"audited_at":"2026-05-19T22:41:39.761878Z","detected_doi":"10.1109/tlt.2023.3323123","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null},{"note":"Identifier '10.1109/tnnls.2023.3242933' is syntactically valid but the DOI registry (doi.org) returned 404, and Crossref / OpenAlex / internal corpus also have no record. The cited work could not be located through any authoritative source.","detector":"doi_compliance","severity":"critical","ref_index":26,"audited_at":"2026-05-19T22:41:39.761878Z","detected_doi":"10.1109/tnnls.2023.3242933","finding_type":"unresolvable_identifier","verdict_class":"cross_source","detected_arxiv_id":null}],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:01:19.554130Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:41:39.761878Z","status":"completed","version":"1.0.0","findings_count":5},{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.699643Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.655894Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"2ffcb0569e6b8a57ff83f2f50c118b97ffe96c7f49d57b46e8e6257a39db037f"},"references":{"count":50,"sample":[{"doi":"10.1111/bjet.12987","year":2020,"title":"Controlled evaluation of a multimodal system to improve oral presentation skills in a real learning setting,","work_id":"9e060509-922e-4f80-882f-640c2213c1eb","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.18608/jla.2024.8411","year":2024,"title":"OpenOPAF: An open source multimodal system for automated feedback for oral presentations,","work_id":"08d09d30-7f23-486b-a50b-1f4e571860ae","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.18178/ijiet.2023.13.5.1879","year":2023,"title":"Evaluation of presentation skills in the context of online learning: A literature review,","work_id":"9354fddf-f3a2-470d-bb4f-51b22fb81026","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1007/s10639-024-13129-","year":2025,"title":"Developing a computer -based tutor utilizing generative artificial intelligence (GAI) and retrieval-augmented generation (RAG),","work_id":"2cf79776-5be5-4f92-b9c4-09081947c7d2","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"10.1109/tlt.2022.3171601","year":2022,"title":"Predicting presentation skill of a speaker using automatic speaker and audience measurement,","work_id":"014af49e-8182-4a9f-b56d-11f9d02b45b5","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":50,"snapshot_sha256":"9b7335dbf9f5ee7d19161136f5ddb30a0035e4487ba6a33c9d5b8099390f0d81","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"686325a7828443e3411afa75753deb4aad5e88d7e38d3b367b4d96a68b078952"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}