Dear Organizers,
We have a question related to GitTables evaluation for CTA. Indeed, the authors of GitTables extracted properties from DBpedia as semantic types, whereas in our work and in the spirit of the CTA task, only classes are considered (properties are only used for CPA task).
Question is: does the Ground Truth consider only DBpedia classes as usual, or does it use a more flexible definition of CTA and includes DBpedia properties?
Thanks in advance.
Regards
Thomas
_________________________________________________________________________________________________________________________ Ce message et ses pieces jointes peuvent contenir des informations confidentielles ou privilegiees et ne doivent donc pas etre diffuses, exploites ou copies sans autorisation. Si vous avez recu ce message par erreur, veuillez le signaler a l'expediteur et le detruire ainsi que les pieces jointes. Les messages electroniques etant susceptibles d'alteration, Orange decline toute responsabilite si ce message a ete altere, deforme ou falsifie. Merci. This message and its attachments may contain confidential or privileged information that may be protected by law; they should not be distributed, used or copied without authorisation. If you have received this email in error, please notify the sender and delete this message and its attachments. As emails may be altered, Orange is not liable for messages that have been modified, changed or falsified. Thank you.
Dear Madelon,
Thanks very much for your answer, that’s clarify indeed.
This dataset is an interesting opportunity to address new challenges, however the target task is not really consistent with the CEA/CTA/CPA definition adopted by the community: using DBpedia properties is closer to CPA, when the Schema ground truth is a mix of CTA/CPA. This may lead to heterogeneous annotations, which might make systems future evaluation much complex.
Anyway this can be an open question, tasks definition may also be challenged if relevant ;)
Regards
Thomas
De : sem-tab-...@googlegroups.com [mailto:sem-tab-...@googlegroups.com]
De la part de Madelon
Envoyé : mercredi 13 octobre 2021 13:54
À : Sem-Tab Challenge
Objet : Re: GiTables CTA
--
You received this message because you are subscribed to the Google Groups "Sem-Tab Challenge" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
sem-tab-challe...@googlegroups.com.
To view this discussion on the web, visit
https://groups.google.com/d/msgid/sem-tab-challenge/10b09b48-716c-4476-b3c0-61a7effbba80n%40googlegroups.com.
On 14 Oct 2021, at 19:47, <thomas...@orange.com> <thomas...@orange.com> wrote:
Dear Madelon,Thanks very much for your answer, that’s clarify indeed.This dataset is an interesting opportunity to address new challenges, however the target task is not really consistent with the CEA/CTA/CPA definition adopted by the community: using DBpedia properties is closer to CPA, when the Schema ground truth is a mix of CTA/CPA. This may lead to heterogeneous annotations, which might make systems future evaluation much complex.Anyway this can be an open question, tasks definition may also be challenged if relevant ;)RegardsThomasDe : sem-tab-...@googlegroups.com [mailto:sem-tab-...@googlegroups.com] De la part de Madelon
Envoyé : mercredi 13 octobre 2021 13:54
À : Sem-Tab Challenge
Objet : Re: GiTables CTADear Thomas,Thanks for participating in this round of SemTab and sharing your question here.
For the GitTables CTA task based on DBpedia, we consider only properties from the DBpedia ontology(http://wikidata.dbpedia.org/services-resources/ontology).
To view this discussion on the web, visit https://groups.google.com/d/msgid/sem-tab-challenge/481_1634237245_61687B3D_481_61_1_9EB5065C625A8644AD4267A2C458DD8D579FF23C%40OPEXCAUBM42.corporate.adroot.infra.ftgroup.