acl-org / acl-anthology

Data and software for building the ACL Anthology.
https://aclanthology.org
Apache License 2.0
438 stars 297 forks source link

Workshop venue field #1117

Open nschneid opened 3 years ago

nschneid commented 3 years ago

Currently the workshops at COLING 2020 list COLING as a venue, e.g.: https://www.aclweb.org/anthology/volumes/2020.dmr-1/

This gives the impression that is a joint event and should be cited as COLING-DMR, which is incorrect.

(Also came up in #1069. Maybe a problem with the ingestion process?)

If we want to list the main conference of a workshop or secondary conference, perhaps this should be a separate field?

nschneid commented 3 years ago

Oh, perhaps I was confusing the case of colocated conferences (like CoNLL) and workshops. I think workshops should be distinguished by including "WS" as a venue, in addition to the main conference, as in https://www.aclweb.org/anthology/volumes/W19-33/

mjpost commented 3 years ago

In the past, the “WS” association happened automatically via the “W” prefix. This is no longer the case, and it is difficult to remember to make this association manually at ingest time. I wonder if we could infer the association: for new style ids, venues not associated with an old style letter could be marked as workshops. Or maybe we could add a “type” field in venues.yaml that defaults to “workshop”.

nschneid commented 3 years ago

What's the status of this? I see none of the ingested 2021 proceedings are marked as WS.

nschneid commented 3 years ago

To the idea of specifying this in venues.yaml, I suspect most venues are (a) always workshops or (b) workshops up to a certain year at which point they "graduate" into conferences (e.g., IWCS became a conference in 2009). So maybe there could be options is_workshop: true and is_workshop_before: 2009.