Output details
11 - Computer Science and Informatics
University of Aberdeen
Output title
Automatically extracting polarity-bearing topics for cross-domain sentiment classification
Type
E - Conference contribution
DOI
-
Name of conference/published proceedings
The 49th Annual Meeting of the Association for Computational Linguistics : Human Language Technologies : Proceedings of the Conference
Volume number
-
Issue number
-
First page of article
123
ISSN of proceedings
-
Year of publication
2011
URL
-
Number of additional authors
2
Additional information
<22> ACL is the most prestigious conference in computational linguistics. This full ACL paper (< 20% acceptance rate) presents a robust and effective framework for sentiment classification by augmenting the original corpus feature space with low dimensional features automatically learned by a topic model. This work, funded by FP7 project ROBUST (http://www.robust-project.eu/), achieved the state-of-the-art accuracy (>90%) in cross-domain sentiment classification and has been used as a standardized framework for sentiment classification within the project.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Citation count
9
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-