SENTIMENT ANALYSIS OF MOBILE REVIEWS USING SENTIWORDNET

Abstract

This paper presents our experimental work on a domain specific feature-based for aspect level sentiment analysis of mobile phone reviews. In decision making, the reviews of others have a important effect on customers to make their choices in regards to do online shopping, choosing products, etc. We devised an rule based domain independent scheme that analyses the textual reviews of a phone and assign a sentiment label on each aspect. The semantic score of subjective sentence is fetched from SentiWordNet Library to calculate their sentiments as +ve, -ve or neutral based on the textual sentence structure. We have used SentiWordNet library as a dataset with two different approaches of selections comprising of adverbs and verbs,adjectives and n-gram feature extraction. We also used our SentiWordNet library to compute the document level sentiment for each phone reviewed and compared its label with results obtained using Alchemy API. The sentiment label of a phone is also compared with the document level sentiment result. The results obtained show that our approach produces a more accurate sentiment label than the simple document level sentiment analysis.