A roadmap of sentiment analysis and its research directions

Sukhnandan Kaur,R. Mohana
DOI: https://doi.org/10.1504/IJKL.2015.073485
2015-12-10
International Journal of Knowledge and Learning
Abstract:The exponential growth of data on the websphere accelerated the need of extracting meaningful information from it. This information can be used for better decision making. The automatic generation of sentiments from the text is called as sentiment analysis (SA). It is a collaborative process of natural language processing and data mining. This paper tries to deeply analyse the existing research work in the area of SA. It presents work done till date and segregates it in terms of level of granularity. An ideal sentiment analyser should have the intellectual capability similar to a human being. This paper mentions a roadmap of the research directions to achieve the goal of ideal sentiment analyser. These research directions include SA based on temporal summarisation, multi-linguality, etc. This paper also mentions various research aspects to work on these research directions. Future research in this direction will also refine the performance of decision making in decision support system.
What problem does this paper attempt to address?