Semantic frame induction through the detection of communities of verbs and their arguments
Abstract Resources such as FrameNet, which provide sets of semantic frame definitions and annotated textual data that maps into the evoked frames, are important for several NLP tasks. However, they are expensive to build and, consequently, are unavailable for many languages and domains. Thus, approa...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
SpringerOpen
2020-09-01
|
Series: | Applied Network Science |
Subjects: | |
Online Access: | http://link.springer.com/article/10.1007/s41109-020-00312-z |
id |
doaj-3f517c48f397486989bf3720d6d03dd7 |
---|---|
record_format |
Article |
spelling |
doaj-3f517c48f397486989bf3720d6d03dd72020-11-25T01:20:43ZengSpringerOpenApplied Network Science2364-82282020-09-015113210.1007/s41109-020-00312-zSemantic frame induction through the detection of communities of verbs and their argumentsEugénio Ribeiro0Andreia Sofia Teixeira1Ricardo Ribeiro2David Martins de Matos3INESC-IDINESC-IDINESC-IDINESC-IDAbstract Resources such as FrameNet, which provide sets of semantic frame definitions and annotated textual data that maps into the evoked frames, are important for several NLP tasks. However, they are expensive to build and, consequently, are unavailable for many languages and domains. Thus, approaches able to induce semantic frames in an unsupervised manner are highly valuable. In this paper we approach that task from a network perspective as a community detection problem that targets the identification of groups of verb instances that evoke the same semantic frame and verb arguments that play the same semantic role. To do so, we apply a graph-clustering algorithm to a graph with contextualized representations of verb instances or arguments as nodes connected by edges if the distance between them is below a threshold that defines the granularity of the induced frames. By applying this approach to the benchmark dataset defined in the context of SemEval 2019, we outperformed all of the previous approaches to the task, achieving the current state-of-the-art performance.http://link.springer.com/article/10.1007/s41109-020-00312-zSemantic framesSemantic rolesContextualized representationsCommunity detectionGraph clustering |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Eugénio Ribeiro Andreia Sofia Teixeira Ricardo Ribeiro David Martins de Matos |
spellingShingle |
Eugénio Ribeiro Andreia Sofia Teixeira Ricardo Ribeiro David Martins de Matos Semantic frame induction through the detection of communities of verbs and their arguments Applied Network Science Semantic frames Semantic roles Contextualized representations Community detection Graph clustering |
author_facet |
Eugénio Ribeiro Andreia Sofia Teixeira Ricardo Ribeiro David Martins de Matos |
author_sort |
Eugénio Ribeiro |
title |
Semantic frame induction through the detection of communities of verbs and their arguments |
title_short |
Semantic frame induction through the detection of communities of verbs and their arguments |
title_full |
Semantic frame induction through the detection of communities of verbs and their arguments |
title_fullStr |
Semantic frame induction through the detection of communities of verbs and their arguments |
title_full_unstemmed |
Semantic frame induction through the detection of communities of verbs and their arguments |
title_sort |
semantic frame induction through the detection of communities of verbs and their arguments |
publisher |
SpringerOpen |
series |
Applied Network Science |
issn |
2364-8228 |
publishDate |
2020-09-01 |
description |
Abstract Resources such as FrameNet, which provide sets of semantic frame definitions and annotated textual data that maps into the evoked frames, are important for several NLP tasks. However, they are expensive to build and, consequently, are unavailable for many languages and domains. Thus, approaches able to induce semantic frames in an unsupervised manner are highly valuable. In this paper we approach that task from a network perspective as a community detection problem that targets the identification of groups of verb instances that evoke the same semantic frame and verb arguments that play the same semantic role. To do so, we apply a graph-clustering algorithm to a graph with contextualized representations of verb instances or arguments as nodes connected by edges if the distance between them is below a threshold that defines the granularity of the induced frames. By applying this approach to the benchmark dataset defined in the context of SemEval 2019, we outperformed all of the previous approaches to the task, achieving the current state-of-the-art performance. |
topic |
Semantic frames Semantic roles Contextualized representations Community detection Graph clustering |
url |
http://link.springer.com/article/10.1007/s41109-020-00312-z |
work_keys_str_mv |
AT eugenioribeiro semanticframeinductionthroughthedetectionofcommunitiesofverbsandtheirarguments AT andreiasofiateixeira semanticframeinductionthroughthedetectionofcommunitiesofverbsandtheirarguments AT ricardoribeiro semanticframeinductionthroughthedetectionofcommunitiesofverbsandtheirarguments AT davidmartinsdematos semanticframeinductionthroughthedetectionofcommunitiesofverbsandtheirarguments |
_version_ |
1725132468615380992 |