Session: Building Recommender Systems: A Case Study with Open Source Software
In this talk, we demonstrate a case study on building a Recommender System using open-source software built as part of the AI Center of Excellence at Fidelity Investments. We focus on fundamental components from feature engineering and selection to model training and testing.
In particular, we show how to generate features from structured and unstructured data via TextWiser and Seq2Pat, how to conduct feature selection via Selective, how to build Bandit-based recommendation algorithms via MABWiser, and finally, how to evaluate performance with respect to recommendation and fairness metrics via Jurity.
This overview can serve as a starting point for software developers and data science practitioners in their efforts in building similar systems.
