Learning greedy policies for the easy-first framework Public Deposited

http://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/hh63sz41k

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Easy-first, a search-based structured prediction approach, has been applied to many NLP tasks including dependency parsing and coreference resolution. This approach employs a learned greedy policy (action scoring function) to make easy decisions first, which constrains the remaining decisions and makes them easier. This thesis studies the problem of learning greedy policies to ensure the success of the Easy-first approach. We propose a novel and principled online learning algorithm that solves an optimization problem to update the weights whenever a mistake happens so that 1) the greedy policy chooses a correct action, and 2) the change to the function is minimized (to avoid over-fitting). The proposed objective is non-convex and optimized via an efficient Concave-Convex Procedure (CCCP). We compare the proposed approach with existing learning approaches on both within-document coreference and cross-document joint entity and event coreference tasks. Results demonstrate that the proposed approach performs better than existing training regimes for Easy-first and is less susceptible to overfitting.
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Keyword
Subject
Rights Statement
Peer Reviewed
Language
Replaces
Additional Information
  • description.provenance : Approved for entry into archive by Laura Wilson(laura.wilson@oregonstate.edu) on 2014-03-20T21:23:16Z (GMT) No. of bitstreams: 1 XieJun2014.pdf: 245676 bytes, checksum: d312aa77758ef66c7dfbf18dc9eebc4c (MD5)
  • description.provenance : Submitted by Jun Xie (xiej@onid.orst.edu) on 2014-03-14T00:50:58Z No. of bitstreams: 1 XieJun2014.pdf: 245676 bytes, checksum: d312aa77758ef66c7dfbf18dc9eebc4c (MD5)
  • description.provenance : Made available in DSpace on 2014-03-20T21:23:16Z (GMT). No. of bitstreams: 1 XieJun2014.pdf: 245676 bytes, checksum: d312aa77758ef66c7dfbf18dc9eebc4c (MD5) Previous issue date: 2014-03-04
  • description.provenance : Approved for entry into archive by Julie Kurtz(julie.kurtz@oregonstate.edu) on 2014-03-14T20:27:42Z (GMT) No. of bitstreams: 1 XieJun2014.pdf: 245676 bytes, checksum: d312aa77758ef66c7dfbf18dc9eebc4c (MD5)

Relationships

In Administrative Set:
Last modified: 08/07/2017

Downloadable Content

Download PDF
Citations:

EndNote | Zotero | Mendeley

Items