BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Relational knowledge in vector spaces - Luis Espinosa-Anke (Univer
 sity of Cardiff)
DTSTART:20190606T100000Z
DTEND:20190606T110000Z
UID:TALK126094@talks.cam.ac.uk
CONTACT:Edoardo Maria Ponti
DESCRIPTION:In this talk a number of unsupervised approaches for learning 
 vectors that capture relational information will be described. The main mo
 tivation behind this is that the amount of information that can be encoded
  in a word embedding is limited\, and constrained by the similarity struct
 ure imposed by the typical methods based on co-occurrence statistics. For 
 example\, the relations holding between lion and zebra\, movie theater and
  popcorn or dog and porch are all seemingly intuitive for us. But it is re
 asonable to assume that an explicit encoding capturing the subtle nature o
 f these relations would be more appropriate than “simply” manipulating
  their word vectors. While such encodings may be acquired from external re
 sources (e.g.\, knowledge bases like ConceptNet or lexical taxonomies like
  WordNet)\, these would be inherently limited\, among others\, by their sy
 mbolic nature. Finally\, in addition to methods for learning relational kn
 owledge\, experimental results will be discussed\, showing their benefit i
 n lexical semantics tasks\, text classification\, and for modeling colloca
 tions.
LOCATION:Board room\, Faculty of English\, 9 West Rd (Sidgwick Site)
END:VEVENT
END:VCALENDAR