Sentiment analysis based on aspect and context fusion using attention encoder with LSTM

Sentiment analysis is a type of natural language processing approach that identifies the emotional tone hidden within a body of text by using machine learning techniques. It would be beneficial to consider the aspects and contexts that are hidden behind the text in order to identify the hidden tone. We propose a hybrid model in this research, based on the fusion of Long short term memory (LSTM) and Encoder with attention for sentiment analysis. We’ve taken into account a variety of factors and different contexts and aspects when analyzing a tweet’s word or phrase. Encoder attention mechanism is used to calculate aspect relevance whereas to obtain the related context, we employ Paragraph2vec to facilitate the process of determining contextual meaning. Paragraph2vec and encoder output features were combined and fed into the LSTM for classification. We carried out experiments using the Twitter Sentiment140 dataset. The results of fusion model is being compared with the LSTM, Bi-directional LSTM and Bidirectional Encoder Representations from Transformers (BERT) model and the results of the tests show that our method outperforms the baseline models that are currently available. © 2022, The Author(s), under exclusive licence to Bharati Vidyapeeth's Institute of Computer Applications and Management.