NYSE Stock Recommender

For my stock trading website tradesimpli.com, I decided there needed to be a better way to improve engagement. Hence I decided to build and integrate a stock recommender engine to recommend similar stocks to the ones that users search for. Initially, all stocks and their corresponding company descriptions were downloaded from the API. NLTK was used to extract keywords from each company’s description. Scikit-Learn was used to build a cosine-similarity matrix for all vectorized keywords, to find most similar stocks to the one inputted. The entire engine was deployed as an API on Microsoft Azure, allowing it to be integrated within the search.
Back to Projects