Loading...
Description
Computer Emergency Response Teams (CERTs) are experts responsible for managing cybersecurity incidents. To identify cyber threats, they consider a wide range of sources from official vulnerability databases to public sources such as Twitter, which has an active cybersecurity community. Due to the high number of topic-related tweets per day, credibility assessment represents an immense effort in the daily work of CERTs. Although approaches for automated credibility assessment have already been developed in previous research, these mainly take peripheral cues into account, although users with domain expertise and a high level of personal involvement also assess content-related cues. We therefore conducted interviews with CERT members to re-evaluate known indicators for automated credibility assessment from an expert perspective. In doing so, we contribute valuable insights to the development of automated approaches for credibility assessment targeting users with high domain knowledge and personal involvement.
Recommended Citation
Basyurt, Ali Sercan; Fromm, Jennifer; Stieglitz, Stefan; and Mirbabaie, Milad, "Credibility of Cyber Threat Communication on Twitter – Expert Evaluation of Indicators for Automated Credibility Assessment" (2022). Wirtschaftsinformatik 2022 Proceedings. 2.
https://aisel.aisnet.org/wi2022/human_rights/human_rights/2
Credibility of Cyber Threat Communication on Twitter – Expert Evaluation of Indicators for Automated Credibility Assessment
Computer Emergency Response Teams (CERTs) are experts responsible for managing cybersecurity incidents. To identify cyber threats, they consider a wide range of sources from official vulnerability databases to public sources such as Twitter, which has an active cybersecurity community. Due to the high number of topic-related tweets per day, credibility assessment represents an immense effort in the daily work of CERTs. Although approaches for automated credibility assessment have already been developed in previous research, these mainly take peripheral cues into account, although users with domain expertise and a high level of personal involvement also assess content-related cues. We therefore conducted interviews with CERT members to re-evaluate known indicators for automated credibility assessment from an expert perspective. In doing so, we contribute valuable insights to the development of automated approaches for credibility assessment targeting users with high domain knowledge and personal involvement.