Obtaining subject data from log files using deep log analysis: case study OhioLINK

Abstract

Traditionally web site statistics and analysis focus on the organization and location information of Internet Protocol addresses and do not analyse sub-network and computer-label information. This paper aims to extract and make use of the information content of sub-network labels in transactional server log files and add an additional level to transaction log analysis. The authors apply microanalytical procedures (i.e. analysis of small segments and sections of log files) to the analysis of log files of the OhioLINK electronic journal service. The authors demonstrate an analysis based on extracted sub-network information and argue that these names can be interpreted as departmental (subject) names. They present an analysis between journal subject groupings and departments based on sub-network labels and find a degree of correlation between department name and subject of journal use. Further, the authors break down journal usage by sub-network label information. The analyses show that sub-network names reflect the physical location of the computer. This presents another possibility of analysing what journals are being used by which academic department.