ExTaSem! Extending, taxonomizing and semantifying domain terminologies
- Submitting institution
-
Cardiff University / Prifysgol Caerdydd
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 102290743
- Type
- E - Conference contribution
- DOI
-
-
- Title of conference / published proceedings
- Thirtieth AAAI Conference on Artificial Intelligence
- First page
- 2594
- Volume
- 0
- Issue
- -
- ISSN
- -
- Open access status
- -
- Month of publication
- February
- Year of publication
- 2016
- URL
-
https://dl.acm.org/doi/10.5555/3016100.3016264
- 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
-
3
- Research group(s)
-
A - Artificial intelligence and data analytics
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Prior to this paper, most work on taxonomy learning was based on co-occurrence statistics in text or external resources, but did not take advantage of the best of both worlds. Here we show ExTaSeM achieved state of the art in the reference datasets available at the time. Our method, based on a two-step process involving disambiguating concepts via a novel strategy based on word embeddings, and CRFs over definition sentences, set the foundations for subsequent works that attempted similar strategies. This work led to successful collaborations with the Sapienza University of Rome (e.g., http://dx.doi.org/10.18653/v1/D16-1041 and http://dx.doi.org/10.18653/v1/S18-1115)
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -