End To End Comments Filtering Feature Using Sentimental Analysis

In the age of digital communication, social media platforms have become essential tools for interaction and information sharing worldwide. However, these platforms often face the challenge of managing negative and harsh comments, which threaten to disrupt a healthy online environment. This paper presents an innovative end-to-end comments filtering feature that uses sentiment analysis to detect and mitigate such undesirable comments. The proposed system employs advanced Multilingual Sentiment Analysis techniques using Long Short-Term Memory (LSTM) networks and the Naive Bayes Classifier to analyze comments in Tamil, Tanglish (a blend of Tamil and English), and English.Once negative comments are identified, the system filters and deletes them, sends warning emails to offending users, and blocks repeat offenders. This approach not only curtails harmful content but also discourages users from posting negative comments in the future. Experimental results demonstrate the system's effectiveness in significantly reducing negative comments and maintaining a positive online community. The combination of LSTM networks and Naive Bayes Classifier provides a robust framework for sentiment analysis and automated moderation, paving the way for healthier social media environments.