AIR: A Semi-Automatic System for Archiving Institutional Repositories

Manual population of institutional repositories with citation data is an extremely timeand resource-consuming process. These costs act as a bottleneck on the fast growth and update of large repositories. This paper aims to describe the AIR system developed at the university of Wolverhampton to address this problem. The system implements a semi-automatic approach for archiving institutional repositories: rstly, it automatically discovers and extracts bibliographical data from the university web site, and, secondly, it interacts with users, authors or librarians, who verify and correct extracted data. The system is integrated with the Wolverhampton Intellectual Repository and E-theses (WIRE), which was designed on the basis of standard software adopted by many UK universities. The evaluation of the AIR system shows its rather high performance and we believe that its exploitation will foster wide dissemination of university research output.