• Lokesh Jain
• Ayush Joshi

This research paper introduces a Twitter Sentiment Analysis System (TSAS) that employs natural language processing (NLP) techniques and machine learning models to analyze and classify tweets based on sentiment. The system leverages Python libraries such as Pandas, NLTK, TextBlob, and Scikit-learn to pre-process and analyze Twitter data. Our research aims to contribute to the field of sentiment analysis on social media platforms, specifically Twitter. The proposed system combines rule-based sentiment analysis using the VADER sentiment analyzer with a logistic regression model trained on labeled data for sentiment classification. The code implements data pre-processing steps, including lowercasing, URL removal, and punctuation removal, to clean tweets. Tokenization, stop word removal, stemming, and sentiment scoring using VADER are employed to enhance the accuracy of sentiment labels. The sentiment of each tweet is classified as positive, negative, or neutral based on predefined thresholds and sentiment probabilities derived from the logistic regression model. Visualization techniques, including Seaborn and Matplotlib, are utilized to display the distribution of sentiment labels and generate word clouds for positive and negative tweets. The system incorporates a graphical user interface (GUI) developed with Tkinter, enabling users to input tweets for real-time sentiment analysis. Additionally, the GUI displays sentiment analysis results, cleaned tweets, and associated emojis. Performance evaluation metrics such as accuracy score, confusion matrices, and visualization of sentiment analysis results are presented. The code also includes features to detect emotions in tweets based on predefined keywords and extract and display the top hashtags.

Sentiment Analysis, Logistic Regression, Social Media

"Tweets Classification Through Natural Language Processing", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.11, Issue 1, page no.e231-e236, January-2024, Available :http://www.jetir.org/papers/JETIR2401427.pdf

"Tweets Classification Through Natural Language Processing", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.11, Issue 1, page no. ppe231-e236, January-2024, Available at : http://www.jetir.org/papers/JETIR2401427.pdf