Show simple item record

dc.contributor.authorCummins, Ronanen
dc.contributor.authorZhang, Mengen
dc.contributor.authorBriscoe, Edwarden
dc.date.accessioned2016-09-27T14:50:18Z
dc.date.available2016-09-27T14:50:18Z
dc.date.issued2016-08-12en
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/260414
dc.description.abstractSupervised machine learning models for automated essay scoring (AES) usually require substantial task-specific training data in order to make accurate predictions for a particular writing task. This limitation hinders their utility, and consequently their deployment in real-world settings. In this paper, we overcome this shortcoming using a constrained multi-task pairwisepreference learning approach that enables the data from multiple tasks to be combined effectively. Furthermore, contrary to some recent research, we show that high performance AES systems can be built with little or no task-specific training data. We perform a detailed study of our approach on a publicly available dataset in scenarios where we have varying amounts of task-specific training data and in scenarios where the number of tasks increases.
dc.languageEnglishen
dc.language.isoenen
dc.publisherAssociation for Computational Linguistics
dc.titleConstrained Multi-Task Learning for Automated Essay Scoringen
dc.typeConference Object
dc.description.versionThis is the author accepted manuscript. The final version is available from Association for Computational Linguistics at http://acl2016.org/index.php?article_id=71.en
prism.publicationDate2016en
prism.publicationNameProceedings of the 54th Annual Meeting of the Association for Computational Linguisticsen
dc.identifier.doi10.17863/CAM.4647
dcterms.dateAccepted2016-05-25en
rioxxterms.versionofrecord10.18653/v1/P16-1075en
rioxxterms.versionNAen
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserveden
rioxxterms.licenseref.startdate2016-08-12en
rioxxterms.typeConference Paper/Proceeding/Abstracten


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record