Semantic Similarity and Selection of Resources Published According to Linked Data Best Practice

The position paper aims at discussing the potential of exploiting linked data best practice to provide metadata documenting domain specific resources created through verbose acquisition-processing pipelines. It argues that resource selection, namely the process engaged to choose a set of resources suitable for a given analysis/design purpose, must be supported by a deep comparison of their metadata. The semantic similarity proposed in our previous works is discussed for this purpose and the main issues to make it scale up to the web of data are introduced. Discussed issues contribute beyond the reengineering of our similarity since they largely apply to every tool which is going to exploit information made available as linked data. A research plan and an exploratory phase facing the presented issues are described remarking the lessons we have learnt so far.