haritha j

DEEP LEARNING RESEARCH INTERN

NIT TRICHY•  May 2023 - July 2023

Developed a hybrid recommendation framework that achieved 0.86 ROC-AUC and 0.972 Precision by integrating Neural Graph Collaborative Filtering (NGCF) and KNN content-based filtering. Evaluated using Spotify dataset of 10,000 user generated playlists that resulted in an input graph of 450,000+ nodes and 100,000,000+ edges