Semantic Analysis on Social Media

Social media sentiment analysis allows to track the opinion about the current ongoing topics and other issues quickly and easily. Several features and techniques for training sentiment classifiers on datasets have been studied in recent years, with inconsistent results. In this paper, we offer a technique for detecting emotion in text and predicting sentiment utilizing semantics as additional features, such as hashtags, which are widely used nowadays in texts. The intelligent technology called as naive bayes is presented in this research for Opinion Mining on social media. It introduces a multinominal naive bayes strategy for extracting semantic and affective components of semi-structured and unstructured text resources from social media sites such as Twitter. This strategy will make the analysis more efficient. This approach will improve the efficiency of the analysis on social media content or the specific data and fuzziness of natural language. The assorted techniques like bag of words, TF-IDF for converting data into features is proposed.