Metoo movement analysis through the lens of social media

Abstract

Sentiment analysis is an errand which is used to analyse people’s opinions which has been derived out of textual data seems productive for palpating various NLP applications. The grievances associated with this task is that, there prevails variety of sentiments within these documents, accompanied with diverse expressions. Therefore, it seems hard to whip out all sentiments employing a dictionary which is commonly used. This work attempts at constructing the domain sentiment dictionary, by employing the external textual data. Besides, various classification models could be utilised to classify the documents congruent to their opinion. We have also implemented topic modelling, emoticon analysis and optimized gender classification in our proposed system. Many sectors have been identified where women are being abused. Clusters are formed for these sectors and the most affected sector is also identified.