Exploring the Semantic Content of Unsupervised Graph Embeddings: An Empirical Study
- Submitting institution
-
University of Newcastle upon Tyne
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 258706-177614-1292
- Type
- D - Journal article
- DOI
-
10.1007/s41019-019-0097-5
- Title of journal
- Data Science and Engineering
- Article number
- -
- First page
- 269
- Volume
- 4
- Issue
- -
- ISSN
- 2364-1185
- Open access status
- Compliant
- Month of publication
- June
- Year of publication
- 2019
- URL
-
https://doi.org/10.1007/s41019-019-0097-5
- 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
-
5
- Research group(s)
-
D - Scalable Computing
- Citation count
- -
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This work explores ways in which interpretability can be brought to unsupervised graph embeddings by assessing which known topological structure is being approximated in the embedding space. Many works have proposed graph embeddings, but this was the first work to assess if these embeddings were learning useful features from the graph. This work has received attention from IBM (Toyotaro Suzumura) and Astra Zeneca, who made a strategic recruitment of one of the authors of the paper based on the work in this paper.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -