Empirical study on the evolution of telecom fraud risks driven by artificial intelligence generated content
The application of knowledge graph and eventic graph technologies in the empirical study of telecom fraud cases driven by artificial intelligence generated content (AIGC) allows for a more intuitive tracing of the evolution of risks during the victimization process, which is of great significance fo...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Beijing Xintong Media Co., Ltd
2025-05-01
|
| Series: | Dianxin kexue |
| Subjects: | |
| Online Access: | http://www.telecomsci.com/thesisDetails#10.11959/j.issn.1000-0801.2025135 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The application of knowledge graph and eventic graph technologies in the empirical study of telecom fraud cases driven by artificial intelligence generated content (AIGC) allows for a more intuitive tracing of the evolution of risks during the victimization process, which is of great significance for countering and early warning of new types of telecom fraud. Based on data from telecom fraud cases implemented using AIGC, semantic role labeling and dependency parsing were firstly performed during data preprocessing. Then, event element recognition and event relationship extraction were constructed to construct knowledge graphs and eventic graphs of the cases. Finally, the key stages and patterns of the evolution of telecom fraud risks were analyzed by combining mathematical statistics with graph technologies. The research revealed that, suspects using AIGC were able to more effectively exploit the phenomenon of confirmation bias to gain the victim’s trust. The evolution patterns of telecom fraud risks driven by AIGC were categorized into three types: the long-chain type evolution pattern systematically identifies complete risk events and their inter-event evolutionary trajectories within cases, while investigation of the star-shaped and composite type evolution patterns enable recognition of divergent risk behavioral patterns and localization of core risk event nodes across homogeneous case clusters, there by establishing a theoretical foundation for developing scientifically rational governance strategies in telecom fraud countermeasures. |
|---|---|
| ISSN: | 1000-0801 |