BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Neural Models for Information Retrieval - Bhaskar Mitra\, Microsof
 t Research Cambridge
DTSTART:20171013T110000Z
DTEND:20171013T120000Z
UID:TALK87641@talks.cam.ac.uk
CONTACT:Amandla Mabona
DESCRIPTION:In the last few years\, neural representation learning approac
 hes have achieved very good performance on many natural language processin
 g (NLP) tasks\, such as language modelling and machine translation. This s
 uggests that neural models will also yield significant performance improve
 ments on information retrieval (IR) tasks\, such as relevance ranking\, ad
 dressing the query-document vocabulary mismatch problem by using semantic 
 rather than lexical matching. IR tasks\, however\, are fundamentally diffe
 rent from NLP tasks leading to new challenges and opportunities for existi
 ng neural representation learning approaches for text.\n \nWe begin this t
 alk with a discussion on text embedding spaces for modelling different typ
 es of relationships between items which makes them suitable for different 
 IR tasks. Next\, we present how topic-specific representations can be more
  effective than learning global embeddings. Finally\, we conclude with an 
 emphasis on dealing with rare terms and concepts for IR\, and how embeddin
 g based approaches can be augmented with neural models for lexical matchin
 g for better retrieval performance. While our discussions are grounded in 
 IR tasks\, the findings and the insights covered during this talk should b
 e generally applicable to other NLP and machine learning tasks.\n \nBio: B
 haskar Mitra is a Principal Applied Scientist at Bing in Microsoft Researc
 h Cambridge. He started at Bing in 2007 (then called Live Search) working 
 on a number of problems related to document ranking\, query formulation\, 
 entity ranking\, and evaluation. His current research interests include re
 presentation learning and neural networks\, and their applications to IR. 
 He co-organized multiple workshops (at SIGIR 2016 and 2017) and tutorials 
 (at WSDM2017 and SIGIR 2017) on neural IR\, and served as a guest editor f
 or the special issue of the Information Retrieval Journal on the same topi
 c. He is currently also pursuing a doctorate at University College London 
 under the supervision of Dr. Emine Yilmaz and Dr. David Barber.
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR