Cognitive agents and machine learning by example : representation with conceptual graphs
- Submitting institution
-
University of Durham
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 116730
- Type
- D - Journal article
- DOI
-
10.1111/coin.12167
- Title of journal
- Computational Intelligence
- Article number
- -
- First page
- 603
- Volume
- 34
- Issue
- 2
- ISSN
- 08247935
- Open access status
- Compliant
- Month of publication
- -
- Year of publication
- 2018
- URL
-
https://doi.org/10.1111/coin.12167
- 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
-
1
- Research group(s)
-
A - Innovative Computing
- Citation count
- 1
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This work succeeds for the first time to create domain-agnostic learning with no prior knowledge, in a systematic and comprehensive way, providing a clear theoretical approach, algorithms and evaluations with several (real-life) datasets and approaches (semantic, probabilistic, neural). It is the first to propose temporal-spatial sequence representations instead of feature vectors; offering a real-time, practically applicable solution, pioneering handling big data. The approach also outperforms the closest related work in entity accuracy. This work led to the founding of Mechion, a start-up working on patient observation devices (https://mechion.com).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -