Balancing the Dimensions of User Intent
Trey Grainger • Back to Haystack Europe 2019
The first step in returning relevant search results is successfully interpreting the user’s intent. This requires combining a holistic understanding of your content, your users, and your domain. Traditional keyword search focuses on the content understanding dimension. Knowledge graphs are then typically built and leveraged to represent an understanding of your domain. Finally, Collaborative recommendations and user profile learning are typically the tools of choice for generating and modeling an understanding of the preferences of each user.
While these systems (search, recommendations, and knowledge graphs) are often built and used in isolation, combining them together is the key to truly understanding a user’s query intent. For example, combining traditional keyword search with your knowledge graph leads to semantic search capabilities, and combining traditional keyword search with recommendations leads to personalized search experiences. Combining all of these dimensions together in an appropriately balanced way will ultimately lead to the most accurate interpretation of a user’s query, resulting in a better query to the core search engine and ultimately a better, more relevant search experience.
In this talk, we’ll demonstrate strategies for delivering and combining each of these dimensions of user intent, and we’ll walk through concrete examples of how to balance the nuances of each so that you also don’t over-personalize, over-contextualize, or under appreciate the nuances of your user’s intent.Download the Slides Watch the Video
Trey GraingerChief Algorithms Officer - Lucidworks
Trey is the Chief Algorithms Officer @ Lucidworks, Author of AI-Powered Search and Solr in Action, Advisor to Presearch and the Southern Data Science Conference, Founder @ Celiaccess, and Investor in multiple startups (primarily tech companies focusing on blockchain, artificial intelligence, healthcare, and education). He is a frequent Public Speaker on Search and Data Science at numerous conferences around the world. He is also a Frequently Published Researcher, with years of research papers and journal articles across the spectrum of search & information retrieval, recommendation systems, data science, and analytics.