Jul 2026· International Seminar on Intelligent Technology and Its Applications· pp. 994-999· 0 citations· 19 references
Abstract
This study compares the performance between traditional feature-based classification methods and transformer architectures in mapping stress, anxiety, and depression conditions in Indonesian-language mental health discourse. The task is formulated as a multi-class classification problem, where each consultation is assigned a single dominant mental health category. By implementing an integrated experimental framework on an online consultation dataset, we tested Gradient Boosting as the baseline model against two specific transformer models, namely IndoBERT and IndoRoBERTa. Experimental findings indicate that transformer-based models consistently outperform traditional approaches, with IndoRoBERTa achieving the highest accuracy of 82%. These results affirm the capability of contextual language representation in capturing complex semantic and linguistic nuances in mental health texts. Nevertheless, this study notes ongoing challenges in differentiating categories with strong semantic overlap, particularly between stress and anxiety symptoms.
It is shown that con-textual transformer/LLM models yield more reliable macro-level performance under imbalance than TF–IDF baselines, particularly for semantically adja- cent classes.
Nehalkumar P. Shah, Mehul P. Barot· International journal of com...· 0 citations
Introduction: Online cancer support forums contain naturalistic accounts of fear, uncertainty, coping, and caregiver strain. Such text may contribute to digital phenotyping as one component of longitudinal, human-supervised monitoring, but the operational link between psychosocial distress and sentiment labels requires explicit evaluation. Methods: We benchmarked representative classical, recurrent, and transformer-based models for four-class ordinal sentiment classification using the Mental Health Insights—Vulnerable Cancer Patients dataset (N = 10,392). Models were evaluated under a single validation-guided 60/20/20 holdout split using weighted F1, macro one-vs-rest AUC, class-specific performance, and paired comparisons. Results: Transformer models achieved the strongest overall performance. ALBERT produced the highest weighted F1 and macro AUC (0.7667 and 0.931, respectively), while BioBERT was closely comparable (weighted F1 = 0.7613; macro AUC = 0.917) and showed slightly higher recall for the “very negative” class (0.8019 vs. 0.7736). Error analysis showed that transformer errors concentrated around ordinal decision boundaries, while residual positive-class errors remained operationally important for supportive workflows. Discussion: These split-specific findings support transformer fine-tuning as a decision-support component for vulnerability-oriented monitoring, while emphasizing calibration, transparent error review, and human oversight rather than autonomous clinical assessment.
Zhongyan Wang, Yuchen Cao, Shuo Xu et al.· Frontiers in Digital Health· 0 citations
This paper presents an experimental study of contextual transformer-based architectures for depression detection from conversations on social media data in the eRisk 2025 dataset by the CLEF Lab. The study addresses two tasks: (1) Depressive Symptom Relevance Detection from conversational posts and (2) Multilingual Depression risk classification. For the first task, different approaches of contextual input configurations (PRE-TEXT-POST) along with domain-specific transformer models (MentalBERT and MentalRoBERTa), probability calibration strategies, a Context-Aware Weighted Fusion (CAWF) mechanism, and multi-seed ensemble methods are thoroughly examined. The results show that MentalBERT, when used with TEXT+POST context, yields the best performance among single models, although the averaging ensemble method leads to further improvement in prediction stability. For the second task, a Hybrid XLM-RoBERTa + CNN + BiLSTM model is proposed for depression detection in multilingual settings with a highly skewed class distribution. The findings show that transformer architectures which are well-calibrated perform better than proposed Context-Aware Weighted Fusion (CAWF) mechanism for Task 1 and balanced evaluation metrics remain important for imbalanced mental-health datasets for Task 2.
Ekta Singh, Jossy P. George, A. Immanuel· International Conference Com...· 0 citations
A hybrid framework for sentiment classification from text, termed QBERT-LSTM, which integrates quantum-enhanced bidirectional encoder representations from transformers (QBERT) with long short-term memory (LSTM) networks, which excel at capturing global context and enhances sequential patterns and temporal features.
Najnin Sultana Shirin, Md. Aminul Islam, Maria Akter Abin et al.· IEEE Access· 0 citations
This research presents a transfer learning approach to analyze the benchmark EmoT corpus of 4401 emotion-labeled Indonesian Tweets, combined with a task-specific data augmentation strategy to enhance model generalization.
Dvir Levi, Phillip M. LaCasse, Lyssa A. White· Discover Artificial Intellig...· 0 citations
It can be concluded that the Bi-LSTM model is effective for automatic text classification of educational questions and has strong potential for further development in technology-based question grouping systems.
Mochamad soffan Muslim, Aviv Yuniar Rahman, R. Pahlevi· Buana Information Technology...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
