Information Retrieval ProjectDescription
- Developed opinion sentiment-aware search engine for Magnificent 7 stocks using web-scrapped Reddit WallStreetBets datasets.
- Devised multi-step classification pipeline incorporating sentiment analysis and sarcasm detection, using state-of-the-art deep-learning and lexicon-based approached with model stacked ensembling.
- Built Apache Solr core with indexing for efficient retrieval and ranking of stock-related posts & sentiment scores.
- Dockerized the application using Docker Compose for reproducibility and scalable deployment.
Skills
