BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:TALK CANCELLED: Cross domain similarities and intra-person changes
  - Professor Maria Liakata - University of Warwick
DTSTART:20210210T150000Z
DTEND:20210210T160000Z
UID:TALK155338@talks.cam.ac.uk
CONTACT:Ben Karniely
DESCRIPTION:I will talk about two conceptually interconnected lines of wor
 k in NLP within my group\; on the one hand identifying semantic similariti
 es between instances (sentences or longer texts but also entities) across 
 domains\, and on the other hand detecting changes within the same person o
 r domain over time. \n\nEven though semantic similarity is a fundamental t
 ask within NLP it can be very challenging when comparisons are made across
  domains as the vocabulary and context can be very different from one doma
 in setting to another. I will talk about recent work of ours where we addr
 ess semantic similarity between two texts in a variety of datasets\, inclu
 ding community question answering\, by injecting domain-specific topic mod
 el information to pre-trained language models [1]. I will also be discussi
 ng how in the case of cross domain entity similarity (and co-reference mor
 e specifically) current models struggle\, some of the reasons behind this 
 and a new resource to help with addressing this problem [2]. The second pa
 rt of my talk can be seen as the flip side of semantic similarity\, where 
 the goal is to look for differences in the representation of the same indi
 vidual (word or person) that signal a change. I will be discussing work of
  ours on sequential modelling of the evolution of a word for semantic chan
 ge detection [3] and how we are developing methods to detect changes in in
 dividuals as part of my UKRI Turing AI fellowship.\n\n[1] Peinelt\, N.\, N
 guyen\, D.\, & Liakata\, M. (2020\, July). tBERT: Topic models and BERT jo
 ining forces for semantic similarity detection. In Proceedings of the 58th
  Annual Meeting of the Association for Computational Linguistics (pp. 7047
 -7055). https://www.aclweb.org/anthology/2020.acl-main.630/\n[2] Ravenscro
 ft\, J.\, Cattan\, A.\, Clare\, A.\, Dagan\, I.\, & Liakata\, M. (2021). C
 D2CR: Co-reference Resolution Across Documents and Domains. arXiv preprint
  arXiv:2101.12637. https://arxiv.org/abs/2101.12637 (Accepted at EACL 2021
 ).\n[3] Tsakalidis\, A.\, & Liakata\, M. (2020\, November). Sequential Mod
 elling of the Evolution of Word Representations for Semantic Change Detect
 ion. In Proceedings of the 2020 Conference on Empirical Methods in Natural
  Language Processing (EMNLP) (pp. 8485-8497). https://www.aclweb.org/antho
 logy/2020.emnlp-main.682.pdf
LOCATION:Online
END:VEVENT
END:VCALENDAR