BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Improving Literature-based Discovery with Neural Networks - Gamal 
 Crichton\, Language Technology Lab
DTSTART:20190307T110000Z
DTEND:20190307T120000Z
UID:TALK121078@talks.cam.ac.uk
CONTACT:Edoardo Maria Ponti
DESCRIPTION:Literature-based Discovery (LBD) uses information from explici
 t statements in literature to generate new knowledge and can thus facilita
 te hypothesis testing and generation from publications to accelerate scien
 tific research. Existing methods\, however\, use methodologies which are i
 nadequate for capturing the complex information available in scientific li
 terature and are prone to proposing spurious or low-quality discoveries. R
 ecent advances in NLP allow for deep textual analysis to obtain a wide cov
 erage of information in text and \nadapt to recognising new entities. Simi
 larly\, recent advances in graph processing have made it possible to do in
 -depth analysis on information represented as graphs to facilitate knowled
 ge discovery. Both advances utilise neural networks extensively. This work
  used neural networks to advance LBD by: improving biomedical NER using mu
 lti-task learning\; \nimproving knowledge discovery from biomedical graphs
  using link prediction\; and improving the ranking of published discoverie
 s by scoring the strength of connection paths. Excitingly\, the latter app
 roaches outperformed those used by the state-of-the-art LION LBD tool. The
 se results show that it is feasible to use neural networks to improve this
  increasingly necessary task and that neural biomedical \nknowledge discov
 ery is potent\, operational and a potentially rich field for further study
 .
LOCATION:Faculty of English\, Room SR24
END:VEVENT
END:VCALENDAR