A Dual-Attention Hierarchical Recurrent Neural Network for Dialogue Act Classification
- Submitting institution
-
University of Aberdeen
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 171386537
- Type
- E - Conference contribution
- DOI
-
10.18653/v1/K19-1036
- Title of conference / published proceedings
- Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL)
- First page
- 383
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- November
- Year of publication
- 2019
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
4
- Research group(s)
-
-
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper proposes a new classification method for dialogue acts using a novel neural network architecture that combines dialogue act and topic information within text representing conversations. The research emerged from a project (EPSRC EP/P011829/1) with a workpackage involving classification of text to improve cybersecurity. In the context of the project, the method was designed to be applicable, in particular, to detection of insider threats from email and other text message data. This paper was accepted in a top-tier Association for Computing Linguistics (ACL) conference with an acceptance rate of 22.66%.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
