Multiword expressions (MWEs) are known as a “pain in the neck” for NLP due to their idiosyncratic behavior. While some categories of MWEs have been addressed by many studies, verbal MWEs (VMWEs), such as to take a decision, to break one’s heart or to turn off, have been rarely modeled. This is notably due to their syntactic variability, which hinders treating them as “words with spaces”. We describe an initiative meant to bring about substantial progress in understanding, modeling and processing VMWEs. It is a joint effort, carried out within a European research network, to elaborate universal terminologies and annotation guidelines for 18 languages. Its main outcome is a multilingual 5-million-word annotated corpus which underlies a shared task on automatic identification of VMWEs.
The PARSEME multilingual corpus of verbal multiword expressions
Johanna MontiMembro del Collaboration Group
;Federico SangatiMembro del Collaboration Group
;
2018-01-01
Abstract
Multiword expressions (MWEs) are known as a “pain in the neck” for NLP due to their idiosyncratic behavior. While some categories of MWEs have been addressed by many studies, verbal MWEs (VMWEs), such as to take a decision, to break one’s heart or to turn off, have been rarely modeled. This is notably due to their syntactic variability, which hinders treating them as “words with spaces”. We describe an initiative meant to bring about substantial progress in understanding, modeling and processing VMWEs. It is a joint effort, carried out within a European research network, to elaborate universal terminologies and annotation guidelines for 18 languages. Its main outcome is a multilingual 5-million-word annotated corpus which underlies a shared task on automatic identification of VMWEs.File | Dimensione | Formato | |
---|---|---|---|
204-3-1319-1-10-20181105.pdf
accesso aperto
Tipologia:
Documento in Post-print
Licenza:
Creative commons
Dimensione
922.05 kB
Formato
Adobe PDF
|
922.05 kB | Adobe PDF | Visualizza/Apri |
parseme_copertina.pdf
accesso aperto
Tipologia:
Documento in Post-print
Licenza:
Creative commons
Dimensione
290.95 kB
Formato
Adobe PDF
|
290.95 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.