Document Clustering Approach for Forensic Analysis: A Survey

Prachi K. Khairkar, D. A. Phalke

Abstract: Digital world is consist of a huge volume of data. In forensic analysis, large amount of data is examined. So its a difficult task for computer examiner to do such analysis in quick period of time. Thats why to do the forensic analysis of documents within short period of time requires special approach such as Document Clustering. This paper reviews different existing Document clustering methods for example, K-means, K medoids, Single Link, Complete Link and Average Link in accordance with computer forensic analysis. In this paper we also give comparative study of different computer forensic analysis techniques and we propose enhance clustering algorithm which will improve accuracy of clustering to finding relevant documents from huge amount of data and speeds up computer examination process.

Keywords: Document clustering, Forensic analysis, Text clustering, Clustering algorithms, Outlier Detection