Skip to content
DraconiticResearch
Search Sources API
197,788 studies31 sourcesUpdated 1 min ago
Peer-reviewedCC BYJournal of Applied Science and Engineering

Towards Intelligent Educational Assessment: A Multimodal NLP Framework for Automated Feedback Generation and Personalized Learning Analytics

  • Jiasheng Ma1
  1. 1Zhengzhou University of Science and Technology, Zhengzhou 450000, China
Read full textDOICite

Abstract

Traditional teaching feedback and student assessment methods often struggle to provide timely, personalized, and comprehensive evaluations in large-scale digital learning environments. To address this core challenge, this study proposes an intelligent teaching feedback and student performance assessment framework based on natural language processing (NLP) and multimodal learning analytics. The framework integrates textual assignments, online discussions, behavioral records, and emotional information to construct a comprehensive learner model. By combining semantic analysis, sentiment analysis, and multimodal data fusion, the proposed approach enables automated performance evaluation and personalized feedback generation. Experimental results from a 16-week controlled trial with 346 students demonstrate that the proposed framework improves student engagement by 88.65%, increases knowledge retention by 65.9%, and achieves a personalized feedback satisfaction rate of 91.3%. In addition, multimodal data fusion lifts comprehensive assessment accuracy from 72% (single-modal text analysis) to 89%, and students’ final exam scores improve by 15% compared with the preintervention baseline. These findings demonstrate the effectiveness of the proposed framework in supporting intelligent educational assessment, personalized learning, and data-driven instructional decision-making.

Related

  1. A Multimodal Engineering GraphRAG Framework for Post-Wind-Event Building Assessment and Recovery Inference
    Seyedeh Fatemeh Mirfakhar, Reda Snaiki, Rocio Lilen Segura
    arXivPreprintCC BY-NC-ND 4.0Full textApplied physics & instrumentationarXiv
  2. Cone-beam CT-based synthetic CT generation for adaptive radiotherapy: methodological paradigms, clinical reliability assessment, and translation
    Yunxian Zhang, Wanting Zhang, Jinbai Huang +4
    Frontiers in OncologyPeer-reviewedCC BY 4.0Full textApplied physics & instrumentationCrossrefEurope PMCPMC
  3. Personalized Modeling of the Diaphragm from Biplanar Radiographs
    Rémi Valentin, William Ndzimbong, Christophe Muth-Seng +5
    Multidisciplinary Biomechanics JournalPeer-reviewedCC BY 4.0Full textGeneral physicsHAL
  4. Factorized Multilinear Codebook Optimization for Large Phased Arrays and Reconfigurable Intelligent Surfaces
    Zhuang Jiapei, Chen Yaocheng
    HALPreprintPublisher termsFull textGeneral physicsHAL
  5. Intelligent intrinsic physical layer security: joint authentication and encryption for data center communications
    Xinshuai Liang, Chongfu Zhang, Yue Lin +3
    Optics ExpressPeer-reviewedPublisher termsOptics & photonicsCrossref

Cite

BibTeXRIS
BibTeX
@article{ma2026towards,
  title = {{Towards Intelligent Educational Assessment: A Multimodal NLP Framework for Automated Feedback Generation and Personalized Learning Analytics}},
  author = {Jiasheng Ma},
  journal = {Journal of Applied Science and Engineering},
  year = {2026},
  month = oct,
  doi = {10.6180/jase.202701_36.037},
  url = {https://doi.org/10.6180/jase.202701_36.037},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Towards Intelligent Educational Assessment: A Multimodal NLP Framework for Automated Feedback Generation and Personalized Learning Analytics
AU  - Jiasheng Ma
PY  - 2026
DA  - 2026/10/31
JO  - Journal of Applied Science and Engineering
DO  - 10.6180/jase.202701_36.037
UR  - https://doi.org/10.6180/jase.202701_36.037
AB  - Traditional teaching feedback and student assessment methods often struggle to provide timely, personalized, and comprehensive evaluations in large-scale digital learning environments. To address this core challenge, this study proposes an intelligent teaching feedback and student performance assessment framework based on natural language processing (NLP) and multimodal learning analytics. The framework integrates textual assignments, online discussions, behavioral records, and emotional information to construct a comprehensive learner model. By combining semantic analysis, sentiment analysis, and multimodal data fusion, the proposed approach enables automated performance evaluation and personalized feedback generation. Experimental results from a 16-week controlled trial with 346 students demonstrate that the proposed framework improves student engagement by 88.65%, increases knowledge retention by 65.9%, and achieves a personalized feedback satisfaction rate of 91.3%. In addition, multimodal data fusion lifts comprehensive assessment accuracy from 72% (single-modal text analysis) to 89%, and students’ final exam scores improve by 15% compared with the preintervention baseline. These findings demonstrate the effectiveness of the proposed framework in supporting intelligent educational assessment, personalized learning, and data-driven instructional decision-making.
SN  - 2708-9967
SN  - 2708-9975
ER  -