Time-Out : Temporal Referencing for Robust Modeling of Lexical Semantic Change

Show full item record



Permalink

http://hdl.handle.net/10138/304380

Citation

Dubossarsky , H , Hengchen , S , Tahmasebi , N & Schlechtweg , D 2019 , Time-Out : Temporal Referencing for Robust Modeling of Lexical Semantic Change . in A Korhonen , D Traum & L Màrquez (eds) , The 57th Annual Meeting of the Association for Computational Linguistics (ACL2019) : Proceedings of the Conference . ACL , Stroudsburg , pp. 457-470 , Annual Meeting of the Association for Computational Linguistics , Florence , Italy , 28/07/2019 . https://doi.org/10.18653/v1/P19-1044

Title: Time-Out : Temporal Referencing for Robust Modeling of Lexical Semantic Change
Author: Dubossarsky, Haim; Hengchen, Simon; Tahmasebi, Nina; Schlechtweg, Dominik
Other contributor: University of Helsinki, Digital Humanities
Korhonen, Anna
Traum, David
Màrquez, Lluís

Publisher: ACL
Date: 2019-07
Language: eng
Number of pages: 14
Belongs to series: The 57th Annual Meeting of the Association for Computational Linguistics (ACL2019) Proceedings of the Conference
ISBN: 978-1-950737-48-2
DOI: https://doi.org/10.18653/v1/P19-1044
URI: http://hdl.handle.net/10138/304380
Abstract: State-of-the-art models of lexical semantic change detection suffer from noise stemming from vector space alignment. We have empirically tested the Temporal Referencing method for lexical semantic change and show that, by avoiding alignment, it is less affected by this noise. We show that, trained on a diachronic corpus, the skip-gram with negative sampling architecture with temporal referencing outperforms alignment models on a synthetic task as well as a manual testset. We introduce a principled way to simulate lexical semantic change and systematically control for possible biases.
Subject: 113 Computer and information sciences
Natural language processing
Computational linguistics
6121 Languages
Rights:


Files in this item

Total number of downloads: Loading...

Files Size Format View
P19_1044_1.pdf 2.562Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record