Extraction2020720phindienglishvegamoviesn Hot ((link)) Official

Report on Unauthorized Distribution (Vegamovies & Similar Sites)

We presented a hybrid keyphrase extraction model tailored for Hindi-English code-mixed text, achieving an F1-score of 0.74. Future work will explore transformer-based models (e.g., IndicBERT) and extend to other Indian languages. Additionally, we plan to integrate the system into a real-time content recommendation pipeline for streaming platforms.