Author

Description

Social media platforms generate extensive user generated content that reflects public opinions, emotions, and experiences. Automatically identifying sentiment from such content is challenging because social media text frequently contains informal expressions, abbreviations, spelling variations, and contextual dependencies. This study proposes a Word2Vec LSTM based sentiment analysis framework for classifying social media content into positive, negative, and neutral sentiments. Initially, the collected social media data are preprocessed through text cleaning, tokenization, stop word removal, and normalization to reduce noise and improve the quality of the input data. Word2Vec is then employed to transform textual words into dense vector representations that capture semantic relationships between words.

Keywords

Social Media, Sentiment Analysis, LSTM