In the current scenario, at the crossroad of computational linguistics and data retrieval opinions and emotions are more valuable than the subject of the document. Linguistic resources are used to retrieve sentiments and also to classify it. Over the internet, not only the large volume of unstructured data is available but also the large amount of text is also generating day by day in the form of blogs, emails, tweets and feedbacks e. t. c. Text analysis is much more mature than unstructured data. Mining is tough for these types of data because of its noisiness and this is the chief bottleneck for designing text mining system. They suffer from spelling mistakes, grammatical errors and improper punctuations because they are informally written. Opinion mining provides a clear platform to catch public's mood by filtering the noise data. It also provides computational techniques used to extract and consolidate individual's opinion from unstructured and noisy text data. This paper tries to cover some techniques and approaches of opinion mining process and also highlight comparative study of some techniques.

Opinion mining Sentiment analysis Social Network Sentiment classification Classification machine learning