How Humans Analyse Lexical Indicators of Sentiments- A Cognitive Analysis Using Reaction-Time

In this paper, we try to understand how the human cognition identifies various sentiments expressed by different lexical indicators of sentiments in opinion sentences. We use the psychological index, Reaction Time (RT) for the analysis of various lexical indicators required for understanding the sentiment polarity. The test bed was developed using linguistic categories of lexical indicators of sentiments and selected sentences which have various levels of sentiments. Experimental results indicate that variations in syntactic categories of the lexical indicators influence the thought in deciding sentiments at varied levels. The results from this work is to be used for fine tuning machine learning algorithms which are used for sentiment analysis and it can also be used in the development of real time applications such as educational tools to better educate students, particularly those with neurocognitive disorders.