You seem define a new XML Schema for your return type in your API. Rather than proliferate the standards we have to support, why not use the CrossRef schema http://www.crossref.org/schema/info/CrossRefSchemaDocumentation4.1.0.pdf, or the dataCite schema, http://schema.datacite.org/, or at least offer data in these formats? For instance, a major feature I'd love to see in ORCID is the ability to add my datasets by DOI and (open access) arXiv preprints and associate them with the publications they refer to, so that they appear as a single "research object". It appears that your data model has no concept of such associations to begin with, as reflected by your XML Schema. Yet this concept is already implemented in the DataCite Schema. Building your internal data model on these existing schemas would both help capture these concepts by leveraging existing work, and provide an API offering data in an already standardized format (which is presumably the format you consume data from CrossRef in already, so this would avoid the data loss of the current conversion...) I do realize that this involves mapping other formats, such as the Scopus data, onto the CrossRef schema rather then your own more loosely defined one, but it seems this could still be worked out automatically, with the result that you would serve much richer data.
Please sign in to leave a comment.