A Prototype Metadata-Driven Decision Support Tool for Scholarly Literature Reviews

Abstract

The aim of this paper is to present the main components of a novel metadata-driven decision support tool for scholarly literature reviews, called PANDORA. The purpose of the tool is to make use of data mining, as a means of identifying semantic similarities between scholarly articles and creating visual models of their relationship. The implementation of user-defined metadata is an important predicate of the tool’s design. As opposed to standard information retrieval tools that treat all metadata equally, the present tool allows users to assign degrees of importance to a selected set of metadata, thus producing a more personalised view of users’ needs. The paper concludes with a summary of the next steps involved in the design and evaluation of the tool.