Development of an Advanced Correspondence and Call Analyzer for Digital Forensics Using Neural Networks and Vector Database

2025

Closed access Metadata (CSL-JSON)
DOI 10.1109/snpd65828.2025.11254885
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

In of the increasing complexity of digital forensic investigations due to technological advancements, this study evaluates digital forensic tools and introduces new analytical systems to advance investigative work. The paper compares some existing forensic tools including Oxygen Forensic, Cellebrite UFED and MOBILedit Pro, all of which can extract and analyze communication data: contacts, calls and correspondences from different devices. An examination of the above tools revealed a lack of features needed to address today’s digital forensics challenges, such as: multilingual semantic search of messages with response translation, multilingual search of unsafe content in correspondence data, communicate and activity time frames in call’s data, which were developed and implemented in the module presented in this paper.