Issues and Future Challenges of Sentiment Analysis for Social Networks- A Survey

Abstract

Sentiment Analysis is a approach that uses Natural Language Processing, Machine Learning, and Deep Learning to computationally identify, categorize, classify, and recover human emotions from unstructured text. The most efficient and popular method for learning from and training social media datasets is deep learning. In a range of applications, including audio, picture, and natural language processing, it has been demonstrated to be more successful. A text block is evaluated using the sentiment technique to determine whether it is positive, negative, or neutral. Not all members of the public express their feelings in the same way; as a result, some do so through comments and ratings while others do so through texts that don't reflect the right frame of mind. This evaluation lists the most typical and well-liked sentiment analysis algorithms for social media data. The classification of positive, negative, and neutral reviews using "LSTM, Knearest method, Random Forest, Support Vector Machine, RNN, and MaLSTM" is explored and examined in this work. All of these classification methods were evaluated along with their drawbacks and difficulties.