DASSH Submission to the 2019 Review of the Australian and New Zealand Standard Research Classification.

While DASSH supports the principles outlined in the Discussion Paper in general, members suggested that the task of research classification could be more effectively achieved and efficiently undertaken were it to draw on techniques arising from advances in artificial intelligence, machine learning, and data science. There would appear to be some scope for achieving this, as demonstrated by the ARC’s incorporation of such methods in the exercise of classifying and assigning research proposals to experts in cognisant fields.

DASSH would welcome the consideration of automated classification methods in the scope of this Review but recognises that the design of a fully automated system (with no human intervention) is perhaps overly ambitious at this stage. DASSH recommends that the ARC, ABS, Stats NZ and MBIE refine and improve the classification system through the current exercise with a view to developing an automated classification system in future.

DASSH further recommends that Field of Research (FoR) codes be classified in a two-level hierarchy, that is, two-digit Divisions and four-digit Groups only. The consensus among members is that broader classifications are more practical. Classification at the six-digit Fields can at times be arbitrary

Publication Details
Publication Year: