Beyond the geotag: assessing implicit geoprivacy risks in visual user-generated content
文献类型：期刊论文
| 作者 | Cheng, Shengjia1; Chen, Peng1; Zhu, Longsheng1; Yan, Yu2 |
| 刊名 | INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
 |
| 出版日期 | 2026-02-04 |
| 页码 | 26 |
| 关键词 | Geoprivacy geo-location risk assessment bayesian network volunteered geographic information |
| ISSN号 | 1365-8816 |
| DOI | 10.1080/13658816.2026.2623524 |
| 英文摘要 | User-Generated Content (UGC) on social media has created a wealth of geographic information. However, it also presents a significant geoprivacy challenge. Image geolocation is the task of identifying an image's location. It offers benefits in domains such as disaster response and autonomous driving. Concurrently, it enables the inference of personal locations from implicit geographic cues within the images. Thus, quantitatively assessing the implicit geoprivacy risk of images is becoming increasingly important. To address this problem, this study constructs a Bayesian Network that includes (1) a comprehensive feature set of 11 metrics in four categories of an image's geolocation, (2) a taxonomy of five core geolocation methods, modeling the causal pathways from the features to the methods, and finally to the vulnerability. We constructed a dataset with 42,576 real-world examples of image geolocation. The model's AUC reaches 0.90 on the test dataset (where 1.0 is perfect and 0.5 is random chance). The results also support that task attractiveness, scene complexity, and landmark level are vital factors in determining the final localization probability. This study provides granular insights into how specific features influence the choice of localization pathways. |
| WOS研究方向 | Computer Science ; Geography ; Physical Geography ; Information Science & Library Science |
| 语种 | 英语 |
| WOS记录号 | WOS:001681541600001 |
| 出版者 | TAYLOR & FRANCIS LTD |
| 源URL | [http://119.78.100.204/handle/2XEOYT63/42811]  |
| 专题 | 中国科学院计算技术研究所 |
| 通讯作者 | Chen, Peng |
| 作者单位 | 1.Peoples Publ Secur Univ China, Sch Informat & Cyber Secur, Beijing, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Beijing, Peoples R China |
| 推荐引用方式 GB/T 7714 | Cheng, Shengjia,Chen, Peng,Zhu, Longsheng,et al. Beyond the geotag: assessing implicit geoprivacy risks in visual user-generated content[J]. INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE,2026:26. |
| APA | Cheng, Shengjia,Chen, Peng,Zhu, Longsheng,&Yan, Yu.(2026).Beyond the geotag: assessing implicit geoprivacy risks in visual user-generated content.INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE,26. |
| MLA | Cheng, Shengjia,et al."Beyond the geotag: assessing implicit geoprivacy risks in visual user-generated content".INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE (2026):26. |
入库方式： OAI收割
来源：计算技术研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。