Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{185433,
        author = {Jenil Girish Rathod and Saloni Mahesh Naik and Aditya Sunil Nalla and Manas Ballal Joshi and Prof. Aparna V. Mote},
        title = {Transformer Models in Digital Journaling: A Review on Mood Detection and Personalization},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {5},
        pages = {1365-1373},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=185433},
        abstract = {Conventional journaling, while useful for self-reflection, tends to be a passive process, falling short of the analysis and feedback required for users to recognize emotional patterns or achieve profound self-awareness [23]. Although recent gains in Natural Language Processing (NLP) have presented solutions, the generalizability and efficacy of text-based emotion detection remain major areas of investigation [4]. This work presents Pocket Journal, a novel, mobile-oriented Android app that seeks to make journaling an engaged, perceptual activity through the use of an advanced pipeline of Transformer-style models. The framework leverages a fine-tuned RoBERTa model for sophisticated mood prediction [1], a fine-tuned BART model for abstractive summarization [14], and incorporates a robust generative LLM (Gemini) to give users on-demand, in-depth behavioral analysis. Main issues in existing text-emotion systems are limited generalizability to different models of emotion [4], reasoning challenges in handling informal or noisy text encountered in actual entries [3], and the propensity of available recommender systems to be shallow, cross-sectional analysis instead of surface-level recommendations [16], [17]. Pocket Journal resolves this through the integration of its multi-stage AI pipeline with personalized media suggestions from third-party APIs (e.g., Spotify, TMDB), rendering a proactive buddy for the improvement of mental health and the promotion of personal development.},
        keywords = {Mental Wellness, Self-Reflection, Artificial Intelligence, Transformer Models, RoBERTa, BART, Natural Language Processing (NLP), Generative LLM},
        month = {October},
        }