A gated multi-scale attention framework for automated short answer grading in Turkish

Basit Öğe Kaydı
dc.contributor.authorKaya, Mustafa
dc.contributor.authorKavun, Mahir
dc.date.accessioned2026-09-01T15:52:38Z
dc.date.available2026-09-01T15:52:38Z
dc.date.issued2026
dc.departmentBayburt Üniversitesi
dc.description.abstractThis study proposes a new deep learning architecture for automated short-answer grading (ASAG) in Turkish. The model integrates transformer-based encoders (BERTurk, RoBERTa-Turkish (RoBERTa), ELECTRA-Turkish (ELECTRA)) with a Gated Multi-Scale Head Attention (GMHA) layer. Additionally, a BiLSTM network and parallel CNN layers are employed to capture both global dependencies and local linguistic patterns. The model was evaluated on a newly created Turkish dataset containing 2200 student responses from 220 learners, each graded by four experts. Among the pretrained Turkish encoders, the BERT-based model with GMHA achieved the best performance (QWK = 0.839, Pearson = 0.856, RMSE = 0.918). The model was also compared with traditional machine learning approaches and large language models (LLMs) using the Leave-One-Question-Out (LOQO) validation protocol. Under this setting, it achieved QWK = 0.337, Pearson = 0.468, and RMSE = 1.213, outperforming all baselines. Ablation studies confirmed the effectiveness of the GMHA and parallel CNN components. These results highlight both the theoretical contribution of multi-scale gated attention and the practical potential of developing reliable automated grading systems for morphologically rich, low-resource languages like Turkish.
dc.identifier.doi10.1016/j.ipm.2026.104791
dc.identifier.issn0306-4573
dc.identifier.issn1873-5371
dc.identifier.issue7
dc.identifier.scopus2-s2.0-105035239903
dc.identifier.scopusqualityQ1
dc.identifier.urihttp://dx.doi.org/10.1016/j.ipm.2026.104791
dc.identifier.urihttps://hdl.handle.net/20.500.12403/8525
dc.identifier.volume63
dc.identifier.wosWOS:001743136400001
dc.identifier.wosqualityQ1
dc.indekslendigikaynakWeb of Science
dc.indekslendigikaynakScopus
dc.language.isoen
dc.publisherElsevier Sci Ltd
dc.relation.ispartofInformation Processing & Management
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanı
dc.rightsinfo:eu-repo/semantics/closedAccess
dc.snmzKA_WOS_20260820
dc.subjectAutomatic Short Answer Grading
dc.subjectTransformer Models
dc.subjectGated Multi-Scale Head Attention
dc.subjectBilstm
dc.subjectCnn
dc.titleA gated multi-scale attention framework for automated short answer grading in Turkish
dc.typeArticle

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu