Spatial and Temporal Sentiment Analysis of Twitter data
The public have used Twitter world wide for expressing opinions. This study focuses on spatio-temporal variation of georeferenced Tweets’ sentiment polarity, with a view to understanding how opinions evolve on Twitter over space and time and across communities of users. More specifically, the questi...
| Main Authors: | , |
|---|---|
| Format: | Book Chapter |
| Published: |
London: Ubiquity Press
2016
|
| Online Access: | http://hdl.handle.net/20.500.11937/36187 |
| _version_ | 1848754699382030336 |
|---|---|
| author | Xia, Jianhong (Cecilia) Zhiwen, S. |
| author_facet | Xia, Jianhong (Cecilia) Zhiwen, S. |
| author_sort | Xia, Jianhong (Cecilia) |
| building | Curtin Institutional Repository |
| collection | Online Access |
| description | The public have used Twitter world wide for expressing opinions. This study focuses on spatio-temporal variation of georeferenced Tweets’ sentiment polarity, with a view to understanding how opinions evolve on Twitter over space and time and across communities of users. More specifically, the question this study tested is whether sentiment polarity on Twitter exhibits specific time-location patterns. The aim of the study is to investigate the spatial and temporal distribution of georeferenced Twitter sentiment polarity within the area of 1 km buffer around the Curtin Bentley campus boundary in Perth, Western Australia. Tweets posted in campus were assigned into six spatial zones and four time zones. A sentiment analysis was then conducted for each zone using the sentiment analyser tool in the Starlight Visual Information System software. The Feature Manipulation Engine was employed to convert non-spatial files into spatial and temporal feature class. The spatial and temporal distribution of Twitter sentiment polarity patterns over space and time was mapped using Geographic Information Systems (GIS). Some interesting results were identified. For example, the highest percentage of positive Tweets occurred in the social science area, while science and engineering and dormitory areas had the highest percentage of negative postings. The number of negative Tweets increases in the library and science and engineering areas as the end of the semester approaches, reaching a peak around an exam period, while the percentage of negative Tweets drops at the end of the semester in the entertainment and sport and dormitory area. This study will provide some insights into understanding students and staff ’s sentiment variation on Twitter, which could be useful for university teaching and learning management. |
| first_indexed | 2025-11-14T08:44:34Z |
| format | Book Chapter |
| id | curtin-20.500.11937-36187 |
| institution | Curtin University Malaysia |
| institution_category | Local University |
| last_indexed | 2025-11-14T08:44:34Z |
| publishDate | 2016 |
| publisher | London: Ubiquity Press |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | curtin-20.500.11937-361872017-05-16T00:47:53Z Spatial and Temporal Sentiment Analysis of Twitter data Xia, Jianhong (Cecilia) Zhiwen, S. The public have used Twitter world wide for expressing opinions. This study focuses on spatio-temporal variation of georeferenced Tweets’ sentiment polarity, with a view to understanding how opinions evolve on Twitter over space and time and across communities of users. More specifically, the question this study tested is whether sentiment polarity on Twitter exhibits specific time-location patterns. The aim of the study is to investigate the spatial and temporal distribution of georeferenced Twitter sentiment polarity within the area of 1 km buffer around the Curtin Bentley campus boundary in Perth, Western Australia. Tweets posted in campus were assigned into six spatial zones and four time zones. A sentiment analysis was then conducted for each zone using the sentiment analyser tool in the Starlight Visual Information System software. The Feature Manipulation Engine was employed to convert non-spatial files into spatial and temporal feature class. The spatial and temporal distribution of Twitter sentiment polarity patterns over space and time was mapped using Geographic Information Systems (GIS). Some interesting results were identified. For example, the highest percentage of positive Tweets occurred in the social science area, while science and engineering and dormitory areas had the highest percentage of negative postings. The number of negative Tweets increases in the library and science and engineering areas as the end of the semester approaches, reaching a peak around an exam period, while the percentage of negative Tweets drops at the end of the semester in the entertainment and sport and dormitory area. This study will provide some insights into understanding students and staff ’s sentiment variation on Twitter, which could be useful for university teaching and learning management. 2016 Book Chapter http://hdl.handle.net/20.500.11937/36187 London: Ubiquity Press fulltext |
| spellingShingle | Xia, Jianhong (Cecilia) Zhiwen, S. Spatial and Temporal Sentiment Analysis of Twitter data |
| title | Spatial and Temporal Sentiment Analysis of Twitter data |
| title_full | Spatial and Temporal Sentiment Analysis of Twitter data |
| title_fullStr | Spatial and Temporal Sentiment Analysis of Twitter data |
| title_full_unstemmed | Spatial and Temporal Sentiment Analysis of Twitter data |
| title_short | Spatial and Temporal Sentiment Analysis of Twitter data |
| title_sort | spatial and temporal sentiment analysis of twitter data |
| url | http://hdl.handle.net/20.500.11937/36187 |