Saurav Kumar

Data Scientist Intern, Data Science Department Contributed to the NewsScope Intelligence System, an advanced data science platform for journalism analytics. Developed a real-time misinformation detection engine using Multinomial Naive Bayes to flag misleading content. Implemented semantic embedding-based topic clustering with UMAP and HDBSCAN for trend identification. Built a hybrid NLP pipeline combining NER, VADER, and RoBERTa for media source credibility and bias scoring. Created supervised models using NMF topic modeling to predict reader engagement levels. Integrated analytical components into a microservices-based Power BI dashboard, enhancing editorial decision-making. Technologies: Python, ML, NLP, Power BI.