A framework for informal language: Opinion mining

Blog, Tweets, or any social networking sites are popularly used weblogs where consumers can broadcast their opinions. It has become a common practise among the users to express their feedback about any product or service. This information in-turn has now become a significant component of the global information. The information provided help's any organization to keep track of the user's opinion about the product or the service that they are providing and helps the company to find their product popularity and its rating in contrast to the other competitors. This information is important for any organization to improve their market share. Thus, these opinion rich resources need to be mined so that one can extract useful information out of it. The information written by the users is in the unstructured form. The techniques are then applied to extract the opinion or the sentiments about the product. The mining of such unstructured data gives rise to new challenges that need to be focused. In this paper, we have proposed a framework to mine the opinions expressed in form of unstructured data. The unstructured data is generally written in the abbreviated form and might not be syntactically correct. Our proposed framework will resolve such challenges which arise due to the commonly used informal language and to handle this we have introduced a new dictionary namely Slang. This dictionary contains the new commonly used abbreviations with their correct matched words and grows with time for every new abbreviated word.