Abstract

With the advent of Web 2.0, Natural Language Processing (NLP) gained attention and a new dimension in the NLP application arena. Sentiment analysis is one such popular modern NLP application that tries to extract the feeling, emotions, opinions, etc., expressed in digital text using NLP techniques. Sentiment analysis has become a subtle application over the last decade and found a firm foundation in AI applications. This is especially applicable to product services, reviews, and recommendations domains. It tries to quantify the sentiment better and helps understand the views expressed in a text leading to better decision-making. There is a variety of approaches available for carrying out sentiment analysis ranging from naive to sophisticated machine learning approaches. However, less attention is being paid to linguistics aspects. We undertook a study to project sentiment analysis concerning linguistic dimensions of natural language text which is the least-explored approach to sentiment analysis. Sentiment components of a text are deeply rooted in its syntactic constituents. The only way to figure out the syntactic constituents is parsing. We have tried to portray the use of dependency parsing for extracting the sentiment components from Hindi input text. We have explored the applications of various dependency relations to derive the possible sentiment compositionality from Hindi sentences. Our study and findings related to sentiment components extraction from dependency parse are presented in this paper with a linguistic insight.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call