Institutional Repositories
Many universities around Australia now have Institutional Repositories (IRs). University IRs are web-accessable repositories of digital research outputs generated by the institution and its members. IRs are generally open and interoperable via protocols such as OAI-PMH and are a means by which institutions manage and provide access to their digital research artifacts and, where appropriate, provide for long-term preservation.

Some of the advantages of IRs, and thus some of the motivations for their establishment include:
  1. Preserving an institution's research outputs
  2. Providing open access to an institution's research outputs
  3. Increasing the visibility and citation impact of the institution’s outputs
  4. Helping to stimulate new forms of scholarly communication.
  5. Serve as tangible indicators of a university's research quality and to demonstrate the scientific, societal, and economic relevance of its research activities. (eg. ERA)

Repository Data
Within Australia, the volume of data collected within university IRs is increasing rapidly. The data contains a great deal of information, mostly bibliographic information encoded using the Dublin Core metadata schema (but not always limited to this schema.)
Examples of available information include,
  • author names (dc:author)
  • journal paper titles (dc: title)
  • abstracts (dc:description)
  • publisher names (dc:publisher)
  • dates (dc:date)
  • full text papers, etc.

As IRs mature it is expected that they will provide open access to many types of data and metadata relating to research outputs.

Data Quality
The ability to extract useful information from the data contained within IRs is strongly influenced by its degree of compliance with accepted standards and how well it is maintained. Some of the data quality issues that we have encountered during the course of this research include,
  1. scarcity of records (this is changing however): some institutions still have very few publicly accessible records.
  2. author ambiguity: no identifier that uniquely idenfies a contributor in the IR.
  3. lack of, or use of non-standard subject codes.
  4. multiple date fields without indication of their meaning.
  5. lack of abstracts (dc:description).

Still, much useful information is presently contained (and is growing) within Australian university IRs and provide interesting insights into research themes and connections within the Australian research environment.