CLASSIFICATION TECHNIQUES OF SENTIMENTS AT DIFFERENT ANALYSIS LEVELS IN CONTEXT OF OPINION EXTRACTION: A REVIEW
In recent days, web marketing and digital environment is pervading everywhere for fulfilling the predominant needs of people especially in Covid-19 pandemic. In general, human beings are strongly associated with sentiments and others opinion. Sentiment analysis contributes a major role in mining the opinions and views of the people. Opinion extraction facilitates the ranking process of search results in an optimal manner. Opinion categorization is done mainly using polarity strength that is obtained through sentimental analysis. This necessitates the implementation of sentiment analysis techniques at Document Level, Sentence Level and Aspect Level. Document level based sentimental analysis calculates the opinion dimensional strength by analyzing the whole document. However, opinion strength of intended domain of the search is determined by applying Sentence level based sentimental analysis. At intrinsic level, to achieve opinion strength of the features or aspects, Aspect level based sentimental analysis is devised. Sentimental techniques are expended in classifying the intentions of users at all levels that facilitates ranking process. This paper examines the classification techniques in all analysis levels. Also summarizes the research strategies enforced for classification, their strengths and captivates the scope of development.