Gagan Jha

Built an end-to-end NLP-based ML system to automatically classify news articles as Fake or Real, addressing the challenge of large-scale misinformation detection. What I did Cleaned and preprocessed unstructured text using NLP techniques (stopwords removal, lemmatization) Converted text into numerical features using TF-IDF vectorization Trained and evaluated a Support Vector Machine (LinearSVC) classifier Performed analysis and experimentation in Kaggle Notebooks Visualized text patterns using word clouds and plots Key Results Handled large, noisy text datasets efficiently Achieved ~99% accuracy on test data Tech Stack Python • Pandas • NumPy • Scikit-learn • NLTK • SpaCy • TF-IDF • SVM • Matplotlib • Seaborn

End-to-end Data Analytics & BI project on real-world Netflix data (~10K records) to uncover trends and build interactive dashboards. What I did Cleaned and preprocessed large datasets using Python (Pandas, NumPy) Performed EDA to analyze genres, popularity, and release trends Created insightful visualizations using Matplotlib & Seaborn Built an interactive Power BI dashboard with KPIs, slicers, and drill-downs Modeled data using Power Query for scalable reporting Key Insights Top 10 most popular movies Top 5 performing genres Peak year for movie releases

NLP and machine learning project using Python, Scikit-learn, NLTK, SpaCy, TF-IDF, and LinearSVC. Cleaned and processed more than 50,000 news articles, benchmarked four classifiers, and achieved 89% test accuracy with LinearSVC. Identified discriminating n-grams for misinformation content moderation.

Data analytics and business intelligence project using Python, Pandas, Seaborn, SQL, and Power BI. Cleaned approximately 8,800 records, handled missing director and cast values through logical imputation, analyzed content trends, and built a multi-page dashboard with genre slicers, year-over-year trends, and regional heat maps.