Assessment of emotional behavior among students through social media identified by Opinion mining

Abstract

One of the web based services, social media provides opportunities for the individual to establish a profile of public or semi-public through limited system and further the other listed for sharing a connection lists that generated from others over the method. According to the web technology evolution, there are internet users accessing large data present over web whereas social media has been generated with huge amount of emotional data in trends of blog posts, tweet, status updates etc. However, the Opinion Mining (OM) is a technique utilized for generating data with useful information about experiencing crowd opinion. Hence, the twitter opinion mining is hard to analyze while comparing with usual opinion mining because of misspelling and word slang present over the contents. Thus, the huge collection of opinions, feedback and views from user’s are accessible across web resources which is most significant for exploring, analyzing and organizing their opinion to assist in making decision. Moreover, the OM is a Natural Language Processing (NLP) and task of extracting information which discovers user’s opinion interpreted in terms of positive or negative comments. Perhaps, the social media has played an important role in cognizing the emotional behavior of university student which may either be accomplished with viable or untenable education. Therefore, this paper has designed a framework by applying opinion mining concept to the student using social media with twitter based content as review dataset which consists of 250 students with their twitter_ID. The unstructured text gets processed and classification results based on emotional polarity to obtain better outcome. Hence, this framework maintain several steps namely opinion extraction, unstructured text processing and polarity classification using NLP. Additionally, the performance of the outcome from the Opinion Mining (OM) gets evaluated using machine learning algorithms.