Applications of Deep Learning to New User Recommendations at Twitter
The cold start problem for new users is a classic challenge for recommender systems. In this talk, I will discuss some deep learning approaches that can be used to address this problem, including using neural networks to train co-embeddings of new users and items, and serving them in an efficient way at runtime via approximate nearest neighbor algorithms like LSH or HNSW. I will also touch on some of the difficulties of evaluating such models both offline and online in the context of A/B tests.
Jay Baxter is a Senior Machine Learning Engineer at Twitter Cortex, where he works on applying scalable machine learning methods to improve Twitter's recommendations and conversational health. Previously, he had worked on a variety of software and machine learning projects, ranging from book search and alerts at Google to entity coreference resolution at Diffeo. He received his M.Eng. and S.B. in Computer Science from MIT, where he led development on a probabilistic database system called BayesDB.