Vol-3Issue-6 2017 IJARIIE-ISSN(O)-2395-4396 6913 www.ijariie.com 150 A Survey On OMCHAR: Opinion-mining Methods and their CHARacteristics Prof. Sandipkumar C. Sagare 1 , Prof. Vinod. G. Khetade 2 , Prof. Mrs . Smita. S. Darbastwar 3 D.K.T.E. Society’s Textile and Engineering Institute, Ichalkaranji 1,2,3 . sagaresandip999@gmail.com 1 , vgkhetade10@gmail.com2 , smitasangewar@gmail.com3 Abstract Natural language processing has an important type called opinion mining that can be used to track the opinion of the public about a particular thing. Opinion mining is widely applicable to context of customer voice such as online surveys, reviews of products and social media. This paper focuses on the different methods and their characteristics for the classification of a given piece of natural language textual data based on the opinions described in it. Keywords ─ Sentiment Analysis, Opinion. 1. INTRODUCTION What others think has always been an important piece of information. For example, “Which smartphone should I buy?”, “Which college should I apply to?”, “Whom should I vote for?”, “Which Professor to work for?”. These examples indicate towards a natural technique of human being to take opinions of others while taking important decisions. Before introduction of WWW, we were taking these opinions manually by asking friends, relatives, experts and consumer reports. Web 2.0 has become advanced so much that many individuals use it to express their opinion and feedback as comments , reviews or question answers on forums, blogs and social websites. This increases the amount of user generated content on the internet. For both user and an organization, this user generated content can be very useful. For example, buyers on online shopping website can check experiences and reviews written by other buyers on that website before purchasing any product. For the online shopping website, these reviews and feedbacks available on the website could be used to make the decisions or make focus groups, surveys in market research. Since very large number of opinions is published on the web, it is difficult for users to analyze all web opinions. To get valuable information from these reviews which are in the form of plain text written in any natural language, we need the help from other domains such as Data Mining and Natural Language Processing (NLP). To summarize and analyze the opinions expressed on websites manually is a difficult task. For that, we require automated sentiment analysis systems [1]. World Wide Web is becoming very popular day by day. So, the task of taking decisions can make use of WWW and analyze the opinions of users. These opinions are helpful for the stakeholders and other public in the decision making. Opinion mining is a technique for retrieving the information through Web blogs, search engines, and social networking sites. Since there is huge number of reviews and opinions in the form of unstructured text, it is not possible to manually summarize this information. So, computational methods that can be efficient are needed for extracting and summarizing the reviews from the web documents [2].