Recognizing Meaning in the Crowd: Building Word Sense Inventories on Amazon Mechanical Turk

DSpace Repository

Show simple item record

dc.contributor.advisor Rumshisky, Anna en_US
dc.contributor.author Botchan, Nicholas
dc.date.accessioned 2012-06-06T19:11:41Z
dc.date.available 2012-06-06T19:11:41Z
dc.date.issued 2012
dc.identifier.uri http://hdl.handle.net/10192/50
dc.description.abstract This thesis explores different strategies for constructing robust, inexpensive and empirically-derived word sense inventories and the corresponding sense-annotated corpus. All strategies explored rely on non-expert linguistic annotations collected through the use of the Amazon Mechanical Turk crowdsourcing marketplace. Experiments using implementation strategies with different quality control mechanisms are reported on in detail. Described herein are multiple best practices discovered through extensive system testing that are required to obtain high quality data given the challenge of using non-expert annotations. Results indicate that it is possible to obtain sense inventories that correlate with the gold standard, extending it in ways that may prove useful in a variety of other Natural Language Processing tasks. en_US
dc.description.sponsorship Brandeis University, Graduate School of Arts and Sciences en_US
dc.format.mimetype application/pdf en_US
dc.language English en_US
dc.language.iso eng en_US
dc.publisher Brandeis University en_US
dc.rights Copyright by Nicholas S. F. Botchan 2012 en_US
dc.subject crowdsourcing en_US
dc.subject WSD en_US
dc.subject word sense en_US
dc.subject inventory en_US
dc.title Recognizing Meaning in the Crowd: Building Word Sense Inventories on Amazon Mechanical Turk en_US
dc.type Thesis en_US
dc.contributor.department Department of Computational Linguistics en_US
dc.degree.name MA en_US
dc.degree.level Masters en_US
dc.degree.discipline Computational Linguistics en_US
dc.degree.grantor Brandeis University, Graduate School of Arts and Sciences en_US


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search BIR


Browse

My Account