Home // International Journal On Advances in Intelligent Systems, volume 7, numbers 1 and 2, 2014 // View article

Towards Establishing an Expert System for Forensic Text Analysis

Authors:
Michael Spranger
Dirk Labudde

Keywords: forensic, ontology, German, text processing, expert system, text analysis, Topic Map, categorization, classification

Abstract:
The analysis of digital media and particularly texts acquired in the context of police securing/seizure is currently a very time-consuming, error-prone and largely manual process. Nevertheless, such analysis are often crucial for finding evidential information in criminal proceedings in general as well as fulfilling any judicial investigation mandate. Therefore, an integrated and knowledge-based computational solution for supporting the analysis and subsequent evaluation process is currently developed by the authors. In this work, we outline the main ideas of this framework and present an approach for categorizing texts with adjustable precision combining rule-based decision formula and machine learning techniques. Furthermore, we introduce a text processing pipeline for deep analysis of forensic texts as well as approaches towards solving domain specific problems like detection and understanding of hidden semantics as well as the automatic assignment of forensic roles.

Pages: 247 to 256

Copyright: Copyright (c) to authors, 2014. Used with permission.

Publication date: June 30, 2014

Published in: journal

ISSN: 1942-2679