Aspect-Based Sentiment Analysis of Marathi Bank Sector Reviews Using Transformer-Based Models

Author(s)	Ms. Simran Najir Maniyar, Dr. Sonali B. Kulkarni, Mr. Gaju S. Chavan
Country	India
Abstract	Aspect-Based Sentiment Analysis (ABSA) has become an important method of fine-grained opinion mining, finding opinion targets and their sentiment polarity. Although much has been done regarding the use of ABSA in English and other high-resource languages, little has been focused on low-resource Indian languages like Marathi, especially in relation to domain-specific applications such as banking. This paper introduces an in-depth transformer-based ABSA of Marathi bank reviews with BERT and its multilingual counterparts. The proposed methodology is based on two steps: aspect term extraction as a sequence labelling task and aspect-level sentiment classification based on contextual sentence-aspect representations. In this work, the proposed aspect-based sentiment analysis system converted 1000 customer reviews into 3115 sentiment instances for each aspect and achieved high classification performance, with a total validation accuracy of 95.18%. Empirical evidence shows that integrating POS-based aspect extraction with transformer-based sentiment classification are effective in fine-grained banking review analysis under low-resource language conditions
Keywords	Aspect-Based Sentiment Analysis, Marathi Language, Banking Reviews, BERT, IndicBERT, Low-Resource NLP
Field	Computer Applications
Published In	Volume 17, Issue 2, April-June 2026
Published On	2026-05-10

About IJSAT Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijsat.org

International Journal on Science and Technology