Thought Vectors, Knowledge Graphs, and Curious Death(?) of Keyword Search
Trey Grainger • Location: Theater 5 • Back to Haystack 2020
The world of information retrieval is changing. BERT, Elmo, and the Sesame Street gang are moving in, shouting the gospel of “thought vectors” as a replacement for traditional keyword search. Meanwhile many search teams are now automatically extracting graph representations of the world, trying their best to also provide more structured answers in the search experience. Poor old keyword search seems so outdated by comparison - is it dead, dying, or simply misunderstood? Contrary to popular belief, embeddings and thought vectors only solve a small subset of search problems, and each of these three tools (keyword search, thought vectors, and knowledge graphs) actually serves a critical role in building the next generation of search experiences. In this talk, we’ll define and highlight the strengths and weaknesses of each of these search methodologies, discuss the role each should play in a modern search solution, and demonstrate where each fails to get the job done and also how they can best complement each other to optimize relevance. We’ll walk through interactive, open source demos showing each of these three types of search in action, demonstrating how to balance the strengths, weaknesses and tradeoffs between them for different user intents and query types.Watch the Video
Trey is the Chief Algorithms Officer at Lucidworks, where he drives vision and practical application of intelligent data science algorithms to power relevant search experiences for hundreds of the worlds biggest and brightest companies. He is also the co-author of Solr in Action, plus more than a dozen additional books, journal articles, and research publications covering industry-leading approaches to semantic search, recommendation systems, and intelligent information retrieval systems. Trey received his Masters in Management of Technology from Georgia Tech, studied Computer Science, Business, and Philosophy at Furman University, and studied Information Retrieval and Web Search at Stanford University.