Order Matters at Fanatics Recommending Sequentially Ordered Products by LSTM Embedded with Word2Vec
This addresses the problem of recommending advanced products to customers in e-commerce, though it appears incremental as it builds on existing Word2Vec and LSTM methods.
The paper tackles the challenge of recommending sequentially ordered products in e-commerce by developing a system that embeds purchased items with Word2Vec and models sequences with stateless LSTM RNN, resulting in improved click-through rates compared to a Word2Vec-only predecessor.
A unique challenge for e-commerce recommendation is that customers are often interested in products that are more advanced than their already purchased products, but not reversed. The few existing recommender systems modeling unidirectional sequence output a limited number of categories or continuous variables. To model the ordered sequence, we design the first recommendation system that both embed purchased items with Word2Vec, and model the sequence with stateless LSTM RNN. The click-through rate of this recommender system in production outperforms its solely Word2Vec based predecessor. Developed in 2017, it was perhaps the first published real-world application that makes distributed predictions of a single machine trained Keras model on Spark slave nodes at a scale of more than 0.4 million columns per row.