The Learning Style Decoder: FSLSM-Guided Behavior Mapping Meets Deep Neural Prediction in LMS Settings
Open accessPreprint — published as 10.3390/computers14090377
Abstract
Personalized learning environments increasingly rely on learner modeling techniques that integrate both explicit and implicit data sources. This study introduces a hybrid profiling methodology that combines psychometric data from an extended Felder–Silverman Learning Style Model questionnaire with behavioral analytics derived from Moodle Learning Management System interaction logs. A structured mapping process is employed to associate over 200 unique log event types with FSLSM cognitive dimensions, enabling dynamic, behavior-driven learner profiles. To evaluate the effectiveness and generalizability of the approach, experiments were conducted across three large-scale datasets: a university dataset from the International Hellenic University, a public dataset from Kaggle, and a combined dataset totaling over 7 million log entries. Deep learning models—including a Sequential Neural Network, BiLSTM, and a pretrained MLSTM-FCN—were trained to predict student performance across regression and classification tasks. Results show high accuracy in binary classification, moderate success in 3-class prediction, and limited effectiveness in fine-grained grade regression and 11-class classification. The proposed framework demonstrates the potential of combining static and behavioral data for scalable personalization in smart learning systems, paving the way for more adaptive, data-driven educational interventions.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Traditional LMS platforms typically follow a one-size-fits-all paradigm that inadequately accommodates individual learner differences.
“traditional LMS platforms typically follow a one-size-fits-all paradigm that inadequately accommodates individual learner differences.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
The Introduction (page not provided) emphasizes individual learner differences and the need for personalized learning. Section 4, “Personalization Process” (page not provided), says most existing personalization approaches are narrow or rely on static profiles. It does not establish that traditional LMS platforms typically deliver the same experience to every learner. (LLM, conf=0.82)
Experimental — accuracy not yet measuredThis mapping translated raw interaction data into cognitively meaningful indicators-for instance, passive forum reading was associated with reflective learning, while repeated quiz attempts were linked to active engagement.
“This mapping translated raw interaction data into cognitively meaningful indicators-for instance, passive forum reading was associated with reflective learning, while repeated quiz attempts were linked to active engagement.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
Section 4.2 says the mapping turns Moodle logs into educational actions and classifies quiz submissions as active. The event-grouping passage (section label “–”; page unavailable) associates forum reading with reflective exploration. Section 5.1 tracks the number of quiz attempts, but describes it as an indicator of consistency and strategy; it does not specifically link repeated attempts to active engagement. (LLM, conf=0.82)
Experimental — accuracy not yet measuredSuch rigidity can result in disengagement and inequitable learning experiences, particularly for learners with diverse cognitive preferences, accessibility needs, or varying levels of prior knowledge.
“Such rigidity can result in disengagement and inequitable learning experiences, particularly for learners with diverse cognitive preferences, accessibility needs, or varying levels of prior knowledge [1,2].”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
The Discussion (section 8; page unavailable) says dynamic learner modeling has advantages over static personalization and can respond to changing engagement and cognitive preferences. It does not establish that rigid approaches cause disengagement or inequitable experiences, particularly for learners with accessibility needs or differing prior knowledge; it says the effects of prior experience have yet to be evaluated. (LLM, conf=0.82)
Experimental — accuracy not yet measuredLearning Management Systems (LMSs) remain largely static and administrative, often failing to support personalization and inclusive access to learning resources.
“Learning Management Systems (LMSs) remain largely static and administrative, often failing to support personalization and inclusive access to learning resources.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
Section 4 (Personalization Process; page unavailable) says many existing approaches rely on static learner profiling and identifies a need for more personalized learning pathways. It does not say that LMSs themselves are largely static or administrative, or that they fail to provide inclusive access to learning resources. (LLM, conf=0.84)
Experimental — accuracy not yet measuredBehavioral indicators such as resource views, quiz attempts, navigation patterns, forum participation, and chatbot interactions are mapped to FSLSM dimensions following established behavior-based profiling methodologies.
“Behavioral indicators-such as resource views, quiz attempts, navigation patterns, forum participation, and chatbot interactions-are mapped to FSLSM dimensions following established behavior-based profiling methodologies [2].”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
Sections 4.2 and 4.4 describe mapping Moodle behaviors—including resource views, quiz activity, forum activity, and content navigation—to FSLSM dimensions using a literature-informed scoring method. The passages do not establish that chatbot interactions were included in that mapping. (LLM, conf=0.88)
Experimental — accuracy not yet measured
Our previous work [2] proposed a hybrid learner profiling methodology that integrates FSLSM questionnaire data with fine-grained Moodle log analytics.
“A representative example is our previous work [2], where we proposed a hybrid learner profiling methodology that integrates FSLSM questionnaire data with fine-grained Moodle log analytics.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
In that study, more than 200 distinct LMS events were systematically mapped to FSLSM dimensions, enabling the construction of dynamic, behavior-informed learner profiles.
“In that study, more than 200 distinct LMS events were systematically mapped to FSLSM dimensions, enabling the construction of dynamic, behavior-informed learner profiles.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
Recent research has shifted toward hybrid and behavior-driven profiling approaches that combine explicit questionnaires with implicit evidence derived from LMS interaction logs.
“In response, recent research has shifted toward hybrid and behavior-driven profiling approaches that combine explicit questionnaires with implicit evidence derived from LMS interaction logs.”
From AI for All: Adaptive, Accessible, and Inclusive Learning Experiences in the Age of Intelligent LMSs (2026) · cites this paper as [2]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
No citation lists for this paper
No citation lists for this paper yet — checked again daily.
References 27 · Cited by 6
Research works printed but not linked: 15
- Hwang, G.J.; Sung, H.Y.; Hung, C.M.; Huang, I. A Learning Style Perspective to Investigate the Necessity of Developing Adaptive Learning Systems. Educational Technology & Society 2013, 16, 188-197.
- Ezzaim, A.; et al. AI-Based Learning Style Detection in Adaptive Learning Systems: A Systematic Literature Review. Journal of Computers in Education 2024.
- Ezzaim, A.; et al. Development, Implementation, and Evaluation of a Machine Learning-Based Multi-Factor Adaptive E-Learning System. IAENG International Journal of Computer Science 2024, 51, 1250-1271.
- Angeioplastis, A.; Aliprantis, J.; Konstantakis, M.; Tsimpiris, A. Predicting Student Performance and Enhancing Learning Outcomes: A Data-Driven Approach Using Educational Data Mining Techniques. Computers 2025, 14, 83. https://doi.org/10.3390/computers14030083. · DOI · Paper page
- Nazempour, R.; Darabi, H. Personalized Learning in Virtual Learning Environments Using Students' Behavior Analysis. Education Sciences 2023, 13, 457.
- Murtaza, M.; Ahmed, Y.; Shamsi, J.A.; Sherwani, F.; Usman, M. AI-Based Personalized E-Learning Systems: Issues, Challenges, and Solutions. IEEE Access 2022, 10, 81323-81347.
- Valencia, Y.; et al. Learning Style Classification by Using Bayesian Networks Based on the Index of Learning Style. Proceedings of ECSEE 2023 2023.
- Buciuman, C.F.; Potra, S. Revolutionizing Education in Industry 4.0: Eye-Tracking and AI for Personalized Learning. Procedia Computer Science 2025, 253, 1658-1667.
- Nafea, S.M.; Siewe, F.; He, Y. On Recommendation of Learning Objects Using Felder-Silverman Learning Style Model. IEEE Access 2020, 8, 125092-125107.
- Fernando-Raguro, M.C.; et al. Extraction of LMS Student Engagement and Behavioral Patterns in Online Education Using Decision Tree and K-Means Algorithm. In Proceedings of the APIT 2022. ACM, 2022, pp. 138-143.
- Hasibuan, M.S.; et al. Recommendation Model for Learning Material Using the Felder-Silverman Learning Style Approach. HighTech and Innovation Journal 2023, 4, 811-822.
- Prabpal, S.; Nitiwatthana, K. Enhancing Learning Style Identification through Advanced Machine Learning Techniques. Sociolytics Journal 2024, 1, 21-33.
- Zagulova, D.; et al. Personalized E-Learning: Relation Between Felder-Silverman Model and Academic Performance. Applied Computer Systems 2019, 24, 25-31.
- Kaur, R.; et al. E-Learning Environment Based Intelligent Profiling System for Enhancing User Adaptation. Electronics 2022, 11, 3354.
- Karim, F.; Majumdar, S.; Darabi, H.; Harford, S. Multivariate LSTM-FCNs for time series classification. Neural networks 2019, 116, 237-245.
- Authors
- Angeioplastis, A., Aliprantis, J., Konstantakis, M., Varsamis, D., & Tsimpiris, A.
- Affiliations
- International Hellenic University (GR)
- University of the Aegean (GR)
- Corresponding author
- Konstantakis, Markos
- ORCID
- Venue
- Computers
- Publisher
- MDPI AG
- Published
- 2025-06-12
- Volume
- 14
- Issue
- 9
- Pages
- 377-377
- ISSN
- 2073-431X (ISSN-L)
- OpenAlex
- W4414155557
- Cited by
- 6
- References
- 27
- Citations by year
- 2026: 6
- FWCI
- 5.34
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- Computers (publishedVersion, cc-by)
- Preprints.org (acceptedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- European Commission
- Erasmus+
- Topics
- Physical Sciences › Computer Science › Computer Science Applications › Online Learning and Analytics
- Physical Sciences › Computer Science › Artificial Intelligence › Intelligent Tutoring Systems and Adaptive Learning
- Social Sciences › Psychology › Developmental and Educational Psychology › Learning Styles and Cognitive Differences
- Keywords
- Felder-Silverman learning styles, student modeling, Moodle, behavioral mapping, student performance prediction, personalized learning, deep neural network, BiLSTM
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
