Aspect-Based Sentiment Analysis for Sinhala Online Reviews (PGDip Defence)

Aspect-Based Sentiment Analysis (ABSA) allows sentiment mining to be performed on a fine-grained level to identify sentiments on particular aspects within a text. This literature review surveys the available methods, data, and models of ABSA, including low-resource applications and the suitability of these approaches to Sinhala online reviews. The review identifies the existing issues, including the lack of data, language complexity, and the lack of resources that are specific to the Sinhala language and explains the recent developments, including neural, transformer-based, and generative frameworks applicable in Sinhala ABSA studies.