Quantum-Inspired Interactive Networks for Conversational Sentiment Analysis.
- Submitting institution
-
The Open University
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 1664457
- Type
- E - Conference contribution
- DOI
-
10.24963/ijcai.2019/755
- Title of conference / published proceedings
- 28th International Joint Conference on Artificial Intelligence (IJCAI2019)
- First page
- 5436
- Volume
- -
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- August
- Year of publication
- 2019
- URL
-
https://www.ijcai19.org
- 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, published at IJCAI 2019, leverages the mathematical formalisms of quantum theory into the long short-term memory network, to learn dynamic inter-speaker influences for conversational sentiment analysis. This was the first quantum-like framework for modelling interactive dialogues, and significantly outperformed the start-of-the-art models from Carnegie Mellon University (Poria et al., 2017). The work is included as recent advances in conversational emotion recognition in (Poria et al., 2019), one of the representative strategies for model contextual relationships amongst utterances in (Hazarika et al., 2020), and also used as a state-of-the-art comparator in recent work from IBM (Chapuis et al., 2020).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -