Sentiment Analysis: Opinion Mining

2691 Words11 Pages

ABSTRACT The term ‘Opinion Mining’ commonly known as Sentiment Analysis, is the branch of study which can be used to analyze and know the sentiments, reactions, attitudes, emotions and opinions of people towards an entity such as different products, services, people, events, topics etc. It is basically the analysis of emotions, effects, subjectivity and the extraction of opinion. So it proves to be a very useful tool for tracking the views and mood of the public about a particular product in the market. The first step would be to know and understand why opinion mining is so important? So talking about the various applications of opinion mining in commercial areas, social media domain and research areas, the importance of opinion mining increases …show more content…

In the human history, it is the first time that we have a big amount of user-rated in social media. Ignoring such data, much research and other things wouldn't have been possible. Chapter 2 APPLICATIONS 2.1 TYPES OF OPINIONS There are two main types of opinions when it comes to the system of opinion mining. The types focus upon the different views of people about a particular product and how they would react upon a service. Also it gives a classification of how people react to a single product and multiple products at the same time. So to classify opinions, there are 2 types which have been introduced which are, Regular Opinions: Regular opinions are those kind of opinions which express views specifically about a particular product or its aspect. Eg. “Fanta tastes good” shows the optimisitc sentiment on the aspect ‘taste’ of the drink. Comparative Opinions: Comparative Opinions are those kind of opinions which compare the multiple entities based on some common aspects. Eg. “Fanta tastes much better than Pepsi” compares Fanta and Pepsi based on the aspect taste and shows a preference for Fanta over Pepsi …show more content…

It is considered to be a profoundly challenging and famous research problem in field of natural language processing. It involves various levels of analysis which helps us to differentiate between different entities and their behavior. The levels are based on the type of document, what kind of expression the user gives and the sentence classification. There are basically 3 levels of analysis, which can be talked about as, 3.1 Different levels of analysis 1. Document Level The document level assumes that each document existing in that level expresses the different opinions as one single entity. It classifies the opinions as whether the whole document shows a optimistic or a negative opinion. For example, for a specific item review, the system checks if the review shows a optimistic or pessimistic opinion about the product. 2. Sentence Level In the sentence level, the task is divided into the lines and checking if each of the sentence expresses a optimistic or a tie opinion. The main focus is upon dividing the different sentences singularly and working upon them. This concept is precisely related to subjectivity classification which differentiates objective sentences from subjective ones. Here, the subjectivity is not equivalent to the sentiment as objectives lines can imply opinion

Open Document