| Keywords [eng] |
Keywords: semantic network model, digital phenotyping, risk assessment, kinship recognition. Raktažodžiai: semantinio tinklo modelis, skaitmeninis fenotipavimas, rizikos vertinimas, kraujo ryšių atpažinimas. |
| Abstract [eng] |
People are the most vulnerable part of cyberattacks, so companies and individuals should consider threats arising from human traits. Many characteristics can describe a person; therefore it is essential to identify which of them can be identified by computer technology and might impact security. In this work, we form the semantic network that structures the automatically recognized traits. Based on the semantic network, the research conducted in this work showed that family members are characterized not only by physical but also by character traits. The identification of family members with a trained network is also an essential part of assessing cybersecurity risks. Some trends between property correlations were also observed. People with a higher extrovert score, for example, are more agreeable and less neurotic. Lithuanians, who participated in this research, are extroverted, open, quite conscientious, and agreeable; they make decisions depending on the situation and rather not take many risks. Similar trends are observed between both genders in different age groups. Although generalizations are possible for groups of people, individualized cyber risk assessment can highlight characteristics that are not specific to a group and draw a person's attention to its weaknesses. Keywords: semantic network model, digital phenotyping, risk assessment, kinship recognition. Scientific publishing: A. Jurevičienė, A. Brilingaitė, and L. Bukauskas, Digital Human in Cybersecurity Risk Assessment, accepted to HCI 2021: Human-Computer Interaction, publishing Springer Computer Science Proceedings, LNCS series, Washington DC, USA, July 24--29, 16 pages, 2021. |