DuEPublicA: Automated bibliometric reports based on the University Bibliography and external citation data

Abstract

This paper describes a web application to generate bibliometric reports based on the University Bibliography and the Scopus citation database. Our goal is to offer an alternative to easy-to-prepare automated reports from commercial sources. These often suffer from an incomplete coverage of publication types and a difficult attribution to people, institutes and universities. Using our University Bibliography as the source to select relevant publications solves the two problems. As it is a local system, maintained and set up by the library, we can include every publication type we want. As the University Bibliography is linked to the identity management system of the university, it enables an easy selection of publications for people, institutes and the whole university. The program is designed as a web application, which collects publications from the University Bibliography, enriches them with citation data from Scopus and performs three kinds of analyses: 1. A general analysis (number and type of publications, publications per year etc.), 2. A citation analysis (average citations per publication, h-index, uncitedness), and 3. An affiliation analysis (home and partner institutions) We tried to keep the code highly generic, so that the inclusion of other databases (Web of Science, IEEE) or other bibliographies is easily feasible. The application is written in Java and XML and uses XSL transformations and LaTeX to generate bibliometric reports as HTML pages and in pdf format. Warnings and alerts are automatically included if the citation analysis covers only a small fraction of the publications from the University Bibliography. In addition, we describe a small tool that helps to collect author details for an analysis.