Efficient Aspect Finding in Sentiment Analysis using Optimal Binary Search Tree

Abstract

Sentiment Analysis is an interesting research area. It useful for analyze the user comments with the help of natural language processing. While consider a huge data and improper data, the sentiment classification with better accuracy is a very big challenge in sentiment analysis. The social media records relevant to Sentiment analysis is a difficult one to identify with a people’s situation, thoughts, and view in the direction of a certain event, which has a number of applications such as election guess and product assessment. This paper focuses on the basic idea of optimal binary search Tree Technique to find the optimal solution (aspect) from a sentence among the several aspects; specifically, we perform the classification process appropriate to various features, which is used to identify the best one aspect among the several aspect terms in a sentence. In the proposed system, We have introduced optimal binary search Tree aspect level sentiment analysis technique. Through this we are getting best and efficient aspect level solution compare to General Tree. The main novelty of the paper is analyzing how aspect sentiment extracted from each word in a sentence through Dictionary based approach with help of the syntactic Patterns. It provides a well understanding of the appropriate role of every words to obtain optimal Aspect in an efficiency way.