arrow_back Operating data pipeline using Airflow @ Slack
Needle in a haystack : entity search on text and graph
Submitted by Uma Sawant (@umasawant) on Tuesday, 1 May 2018
Web search today is moving towards displaying “answers” rather than making the user browse through pages to find what they want. “Entity” search queries, where the expected answer is a list or a set of objects; form more than 40% of today’s Web search. Yet the current approaches for answering such queries are quite brittle.
We improve the state-of-the-art by infusing the semantic information of entities, their types and inter-relations into free text from Web documents. The result is a much more robust QA engine. In this talk, we will discuss the challenges in Web-scale entity search, our approach of combining structured and unstructured information and how that allows us to answer user queries better.
Uma Sawant is an applied reseach engineer working in LinkedIn. She previously acquired her PhD and masters in Computer Science, from IIT Bombay, India. Her PhD thesis centers around Entity search and she has given a number of talks in this field, including international peer-reviewed conferences as well as data meetups.