International Workshop on Semantic Evaluation

SemEval 2020: Program

SemEval-2020 will be colocated with COLING 2020. All times shown are Central European Time (CET, UTC+1)

Proceedings in the ACL Anthology

(schedule updated: 3 Dec)

Saturday, December 12

Session One: Sentiment analysis and societal applications (Tasks 9, 11, 12)

14:00-14:30 Opening remarks, Q&A for oral presentations

14:30-16:00 Poster session

Session Two: *SEM/SemEval keynote talk: Afra Alishahi

16:00-17:00 Grounded language learning, from sounds and images to meaning, Afra Alishahi, University of Tilburg

Abstract: Humans learn to understand speech from weak and noisy supervision: they manage to extract structure and meaning from speech by simply being exposed to utterances situated and grounded in their daily sensory experience. Emulating this remarkable skill has been the goal of numerous studies; however researchers have often used severely simplified settings where either the language input or the extralinguistic sensory input, or both, are small-scale and symbolically represented. I present a series of studies on modelling visually grounded language understanding. Using variations of recurrent neural networks to model the temporal nature of spoken language, we examine how form and meaning-based linguistic knowledge emerges from the input signal.

Session Three: Lexical semantics (Tasks 1, 2, 3)

17:00-17:30 Q&A for oral presentations

17:30-19:00 Poster session

Sunday, December 13

Session Four: Common Sense Knowledge and Reasoning, Knowledge Extraction (Tasks 4, 5, 6)

14:00-14:30 Q&A for oral presentations

14:30-16:00 Poster session

Session Five: SemEval keynote talk: Jackie C.K. Cheung

16:00-17:00 From Discourse Structure to Semantics in Automatic Summarization, Jackie C.K. Cheung, McGill University

Abstract: The stereotypical discourse structure of a genre is often a good indicator of importance for content selection in automatic summarization. For example, the opening sentences of a news article usually form a good summary of it. However, relying on discourse structure could arguably be seen as a crutch on our way towards modelling the semantic content of source documents and the summaries. In this talk, I discuss the possibilities enabled by more explicitly thinking about semantics for neural abstractive summarization, with a focus on datasets and evaluations. I will discuss recent work on detecting and correcting factual inconsistencies in abstractive summaries. I will also emphasize the need for new summarization tasks that target semantic generalization and aggregation.

Session Six: Humour and emphasis (Tasks 7, 8, 10)

17:00-17:30 Q&A for oral presentations

17:30-19:00 Poster session