A Review of Challenges in Aspect-Based Sentiment Analysis Sub-tasks in Resource-scarce Indian Languages

Large-scale research by authors has covered aspect-based sentiment analysis (ABSA) and its sub-tasks, such as Aspect Term Extraction (ATE), Aspect Term Polarity (ATP), Aspect Category Detection (ACD), Aspect Category Polarity (ACP) in resource-rich monolingual English text. In this present paper, we have focused on research progress with respect to language models explored by different researchers, challenges, and research gaps in resource-scarce Indian Languages (ILs). In the proposed survey, we have considered seven ILs, such as Bengali, Hindi, Kannada, Malayalam, Nepali, Telugu, and Urdu, in which the authors have contributed research on ABSA and its sub-tasks in ILs. This survey will guide researchers to know about modeling approaches so far explored by researchers for corpus evaluation of each language in each paper considered and envisaging comparison of the performance of language models along with challenges towards developing language models and research gaps in each language those are identified to motivate researchers as future research scope. The present study will help researchers to explore different domains and dataset statistics of benchmark ABSA dataset in detail for each ILs language considered for the ABSA task and its sub-tasks.