Predicting environmental features by learning spatiotemporal embeddings from social media
- Submitting institution
-
University of the West of England, Bristol
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 5986231
- Type
- D - Journal article
- DOI
-
10.1016/j.ecoinf.2019.101031
- Title of journal
- Ecological Informatics
- Article number
- 101031
- First page
- -
- Volume
- 55
- Issue
- -
- ISSN
- 1574-9541
- Open access status
- Compliant
- 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
-
2
- Research group(s)
-
-
- Citation count
- 2
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper proposes a novel model for learning vector space embeddings of spatiotemporal entities which is able to integrate structured environmental information with textual information from Flickr tags. We also introduce a new smoothing method to deal with the sparsity of Flickr data over space and time. The qualitative and quantitative evaluation of the proposed model has proven to be advantageous compared with baselines that rely only on Flickr or only on traditional sources. The initial experiments and findings were presented in the European Conference on Information Retrieval (ECIR 2019).
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
