CS 201: Yahoo Knowledge: Building a Web-Scale Knowledge Graph, MIKE WELCH, Yahoo!

Speaker: Mike Welch
Affiliation: Yahoo!

WELCH-PIC

 

ABSTRACT: Large scale knowledge graphs play a crucial role in the modern Web search landscape.  A shift from the traditional document-centric “ten blue links” view of the world to one with direct answers, contextual awareness, anticipatory search, and actionable results depends on high coverage and a deep understanding of real world entities and relationships between them.   A single source of data is rarely sufficient to build these richer results for even one domain across all markets.  The Yahoo Knowledge team is responsible for collecting data from heterogeneous sources all across the Web, synthesizing it, and serving it as a unified graph containing billions of interconnected facts.  In this talk I will describe some of the core principles of knowledge graph construction: data acquisition, integration, and data fusion.  I will also discuss how the Yahoo Knowledge Graph platform applies these principles at scale to achieve the breadth and depth required by a Web search engine. BIO: Mike Welch is a Principal Engineer at Yahoo, co-architect and the top code contributor for Yahoo’s Knowledge Graph platforms, and a lifelong Bruin (B.S. ’03, M.S. ’05, Ph.D ’10).  He is the primary inventor on 10 patents and has published his work in multiple peer reviewed conferences including WWW, WSDM, SIGIR, VLDB, and CIKM.  His current work focuses on the combination of science and engineering required to perform entity resolution and data fusion at Web scale.  Prior to joining Yahoo, Mike received his Ph.D in Computer Science from UCLA under the guidance of Professor Junghoo Cho.

 Hosted by Professor Ameet Talwalkar

REFRESHMENTS at 3:45 pm, SPEAKER at 4:15 pm

Date/Time:
Date(s) - Jan 21, 2016
4:15 pm - 5:45 pm

Location:
3400 Boelter Hall
420 Westwood Plaza Los Angeles California 90095