Towards real-time, country-level location classification of worldwide tweets
- Submitting institution
-
University of St Andrews
- Unit of assessment
- 11 - Computer Science and Informatics
- Output identifier
- 262880938
- Type
- D - Journal article
- DOI
-
10.1109/TKDE.2017.2698463
- Title of journal
- IEEE Transactions on Knowledge and Data Engineering
- Article number
- -
- First page
- 2053
- Volume
- 29
- Issue
- 9
- ISSN
- 1041-4347
- Open access status
- Compliant
- Month of publication
- April
- Year of publication
- 2017
- 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
-
5
- Research group(s)
-
B - Systems
- Citation count
- 16
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- Only a small percentage of tweets contains geo-location information as GPS coordinates and user-supplied information is also sparse and potentially unreliable. This poses a significant problem for applications that depend on identification of the geographic origin of tweets such as trending topic detection. Our method for accurate realtime country-level classification of tweets was used in the production of a social media driven program for BBC 5Live https://www.bbc.co.uk/blogs/5live/entries/7e6b5290-541e-35e2-9afe-8f5eb00cb467. Unlike previous work, it achieves real-time operation using only features inherent in a tweet and is not specific to a single country but works on the unfiltered Twitter stream from 217 countries.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
