Activate 2018

This session presents how voice search has enabled fans to discover not only events but also tickets at StubHub. Using AI for entity detection, entity disambiguation, ranking and understanding unstructured data, we are able to provide a rich and powerful experience. Combining homegrown NLP and external NLP libraries and services, we were able to productize SolrCloud and AI capabilities to discover best value tickets for a price range and seat location.

We will go over design and architecture of NLP system and query classifier to identify NLP vs normal search query. Voice search using mobile devices is a fast-growing driver of traffic for stubhub and e-commerce site search. We will share how it is helping fans to find right seats in shortest time and hence improve conversion, when they have a choice of many traditional filters to find tickets by attributes like price, quantity, venue section and seat location for aisle or handicap.