Challenges and Opportunities in Applied Machine Learning for Product Search at Wayfair

Abstract: Product Search in eCommerce presents unique challenges that the traditional information retrieval literature does not address. At Wayfair, we encounter these challenges first hand: handling a variety of customer search goals, working with volatile product catalog, personalizing shopping journeys, etc.  In this talk, we will introduce the deep learning models for Wayfair’s search query understanding and multi-modal product retrieval, which address some of these challenges. Finally, we will discuss open research opportunities in Product Search, and invite the academic community to tackle.

Biography: Weiyi is a Senior Manager at Wayfair, where she leads the Data Science Search team. Wayfair is one of the world’s largest online markets for home goods. Wayfair Data Science Search team creates machine learning solutions to improve the product search experience for tens of millions of customers every day, and automates the end-to-end workflow for building, deploying, and maintaining models at scale. Previously, Weiyi was a Research Science Manager at Nuance, building state-of-the-art abstractive summarization, information extraction and text classification models to improve clinical documentation quality at bedside. Prior to that, Weiyi’s doctoral research focused on temporal reasoning and information extraction in clinical narratives.