Recommendation techniques in forensic data analysis: a new approach

Abstract

Data mining for digital forensic analysis is a branch of Computer Science focused on pattern extraction from largescale data which has been used to support analysts when trying to solve crimes. One of the most promising applications of data mining algorithms is to build recommendation systems, aiming to propose future directions to the investigation and to guide the analyst through the process. In this paper we propose a new approach, architecture and framework with the purpose of taking advantage of the recommender systems techniques to the forensic field and provide examples of their applicability to different use cases involving large scale collections of multimedia information related to a defined forensic case.