Identification of Latent Aspects Using Bayesian Non-Parametric Models

Abstract

Sentiment analysis is increasingly viewed as a vital task both from an academic and a commercial standpoint. The majority of current approaches, however, attempt to detect the overall polarity of a sentence, paragraph, or text span, regardless of the entities (e.g., laptops, restaurants) and their aspects (e.g., battery, screen; food, service) are mentioned. By contrast, this task is concerned with aspect based sentiment analysis (ABSA), where the goal is to identify the aspects of given target entities and the sentiment expressed towards each aspect. Although, it is mandatory to evaluate both the positive and negative reviews of the customer. Since the analysis of negative review will help to improve the business specification. This system is mainly about evaluating reviews for specific entities of any products based on the three slots: aspect categories, opinion target expressions, and polarity classification. Aspect category is to identify an entity E and attribute A pair from which the reviews are expressed. Opinion target expressions are based on customer review for each product. Polarity classification is used to identify or express the positive and negative reviews of the product. Also the project helps to infer the hidden latent topics using a Bayesian Non Parametric model. The existing models failed to provide the Inter-Dependency between aspects based analysis & over-all ratings. Initially this phase is all about finding the aspect based sentiment analysis using Naive Bayes algorithm.