AI-Based Sentiment Analysis on Social Media Using Deep Learning and NLP Techniques

Shraddha Dilip Kapase, Dr. Santosh Gaikwad

Abstract: The current research is dedicated to analyzing the user opinions expressed on social networks using novel computational techniques. With an ever-growing number of online communications, there arises a vast amount of text data, which cannot be processed manually. To address this issue, the proposed solution will apply Natural Language Processing approaches combined with different machine learning and deep learning models to automatically detect the sentiments in the given text. The models used for the study are as follows: SVM, LSTM, CNN + LSTM, and BERT model. They will help to classify social network content into positive, negative, and neutral sentiments. Such studies have practical applications in fields such as brand monitoring, customer feedback analysis, and public opinion tracking. The obtained outcomes indicate that combining classical models with deep learning models improves their efficiency in sentiment classification.

Keywords: Sentiment Analysis, Social Media Data, Natural Language Processing, Deep Learning Models, Opinion Mining