Amanda Hill talking about ‘a new approach to name authority’

The ‘Names’ project – JISC funded project based at Mimas to investigate and build a prototype/pilot system to deal with name authority.

What’s the problem? Amanda talking about how her father bought books from an Amazon Wishlist – thinking it was her’s, but actually turned out to be another Amanda Hill who happened to live in the same town. More serious examples – someone operated on in error because of a name mix up.

Amanda gives a (fictitious) example of how many different variations on names can appear in academic publications. From a real example: [edit: for clarity Amanda gave a fictitious example I missed, and then followed that with a real example which was taken from this blog post, and this real example is the one given here]

• M. Lynne Murphy
• Lynne M. Murphy
• L. M. Murphy

All same person…

So names can be confusing. You may want to retrieve all articles by a particular individual – may have to think of all forms of the name, and exclude people with the same name.

Zetoc is a service from the British Library and Mimas which contains information about journal articles – and so millions of names of active researchers – but not complete names always, because generally initials only used in academic publication.

So used Zetoc to pre-populate the Names system – assign unique identifiers – then start to enhance from other places.

Currently project is testing matching algorithms, reviewing data structure, updating data mappings for other standards (MARC, CERIF, etc.), collaborating with potential data providers and Names users.

You can search the pilot system at http://names.mimas.ac.uk/advanced-search.php Amanda doing demonstration – e.g. Abbott, S – note that the identifiers currently assigned are not persistent – but there will be persistent identifiers once they finished tweaking data/algorithms/etc. Can also see relationships to others (based on co-publication I think) – e.g. http://names.mimas.ac.uk/individual/60767.html?outputFields=collaborativeRelationships

In the long term want to allow researcher to edit their own information (e.g. specified a preferred name).

Names project not the only activity in this area – also:

• ORCID
• VIAF and WorldCat Identities
• ISNI (International Standard Name Identifier)
• Various country-based initiatives

Q: How far are names matched automatically – is there any human intervention

A: Currently completely automated – but checking how accurate this is. Also erring on side of caution – so if any doubt they don’t merge identities. Matching will improve as they get more data.

Q: What about dealing with non-english script names?

A: Names project focussed on the UK research community, but Zetoc data not limited to this, and system designed to cope with different versions of names (and name changes)