Authors: Shubham Rathod, Dr.Rajeshwari Kannan
Abstract: Social media platforms such as Twitter generate huge textual data every day, making sentiment analysis an important research area in Natural Language Processing (NLP). As there are many tweets that are noisy and reqired some contextual background . So the Research proposes the hybrid architecture of CNN and BiGRU so solve such issue. The Dataset which has been used is Sentiment 140 which contains 1.6 million entries.The methodology used is preprocessing , tokenization , CNN feature extraction and BiGRU contextual learning.The result shows the final accuracy using such architecture comes out to be 77.13%.Along with that there is use of flask web applicaton to create an interface.
DOI:
