Author:
Li Gaoxiang,Ye Yuzhong,Shi Yangfei,Chen Jinlin,Chen Dexing,Li Anyang
Abstract
AbstractWith the increasing of the telecom network fraud in China, SMS (Short Message Service) has became an important channel exploited by the criminals to contact victims. Due to the tiny amount compared with normal SMS, the high proportion of malicious adversarial characters, and the lack of knowledge to specific fraud types, it is still challenging to identify the fraud SMS efficiently. In this paper, we firstly conduct a measurement study to explore the characteristics of the fraud SMS. Based on the exploration, we propose a two-stage algorithm called TFC. TFC can quickly filter out normal SMS in the first stage with two indicator functions, and then easily identifies the category of fraud SMS in the second stage by combining the semantic deep features and the domain-knowledge based artificial features. We conduct two real-world SMS datasets for extensive experiments, and the results show that TFC successfully reduces calculation cost and achieves better performance in distinguishing various categories of fraud SMS.
Publisher
Springer Nature Singapore
Reference30 articles.
1. Supreme People's Procuratorate sets up a steering group to crackdown the rapidly spread cybercrime. http://www.xinhuanet.com/legal/2020-04/08/c_1125829699.htm. Accessed 07 Apr 2021
2. The survey report shows that three telecom operators have achieved nearly 100% scam SMS processing in September 2020. https://www.sohu.com/a/430647655_287480. Accessed 07 Apr 2021
3. China public security department cracked 256,000 cases of telecommunications network fraud in 2020. http://www.gov.cn/xinwen/2021-01/02/content_5576223.htm. Accessed 07 Apr 2021
4. Wikipedia, PINYIN. https://en.wikipedia.org/wiki/Pinyin. Accessed 07 Apr 2021
5. Cui, Y., et al.: Revisiting pre-trained models for Chinese natural language processing. In: Proceedings of Conference on Empirical Methods in Natural Language Processing: Findings (2020)