Application of Artificial Intelligence for Forming Evidence Base of Crimes in Smart Homes

Abstract

This paper presents a solution for forming an evidence base for crimes involving IoT devices in smart homes. It introduces artificial intelligence model assistant as the way to identify potential crime evidence. The crime evidence is identified by detecting abnormal behavior related to user and the sequence of data coming into the controller of a smart home [1]. The proposed approach analyzes data from smart home ecosystems and determines the likelihood of specific events being related to a crime [2-3]. The data is analyzed by artificial intelligence model bringing outputs in term of probabilities. The results are presented as percentages, reflecting the probability of connection of specific event and a sequence of events to a crime. Final calculations are brought  to attention of a forensic specialist as a half sum of 2 probabilities. Using provided data the digital forensics criminalist can base further investigation and build theories on circumstances of a crime. This approach helps specialists to concentrate attention on valuable events by highlighting them and reducing the time spent reviewing data.