ABSTRACT

The sentimental analysis is the major concept of the analysis emerged with the advance of social media. Our purpose is to define the sentiment analysis of a Twitter and Facebook comment Positive, Negative and Neutral by using the Vectorization and instance selection approach. In the main sentiment analysis applications using these methods, the sentiment keywords, sentence plays a main role. It is most important to make a sentence and feature based covering various sentiment words. For the reason, we analyze the paper issues how to divide and list-words, sentences present on the focus into binary dictionaries. We have implemented feature extraction method, instance selection and a novel involuntary method to make the positive, negative and neutral dictionaries that search the sentiment keywords present in the comments. More significantly, our idea allows increasing these dictionaries with an en-richest phase. Lastly, by using these prepared datasets, we identify the category of the sentiments of the comments and reviews in social media networks. We consider our method by comparing to human authentication and classification. Our consequences are also real and consistent.