EThOSnet Home

EThOS Toolkit (guidelines & further information available here)

Resources

News

Contact Us

Who's Involved?

Existing Arrangements

The Vision

The 7 Strands

Project Documentation

Theses Digitisation Project

Old EThOS website

Questions and Notes on the NAMES Project Presentation

Q: Are you planning to interact with other NAMES servers around the globe?
A: That’s something that we’ve looked at in our landscape report and in the future that might be the case. At the moment we’re just keeping it local. But that’s not something that we’ve ruled out.

Q: I was wondering if you’re planning to bulk-import any existing sources into this, like the Library of Congress (LoC)?
A: Yes we have actually spoken with the LoC and I think that will be something that happens. At the moment, we’re still feeling our way with what direction we want to take it and the scale of it. More and more we’re realising that we can’t just do this on a national level, it needs to be done on an international level.
Q: I guess with the LoC, humanities might be a strong source but I don’t think it will be a strong source for scientific data.
A: An example, we’re working with UKPMC which deals with more the medical and biological side of things. It was actually one of the big drivers for this project because they wanted to look at how their different databases that deal with grantors and grantees and they wanted to look at the outcomes of grants and be able to identify individuals and match things up. They’ve actually provided some information for testing purposes at the moment, but also for subsequent inclusion in a live service – which hopefully will happen one day.

Q: It sounds quite simple, but how does it make that decision about disambiguation? For example, when you get your information in from three different citations does it know that Joe Bloggs is the same as J Bloggs and also the same as J L Bloggs, but also isn’t the same as J Bloggs who happens to be Jane Bloggs?
A: Obviously that is the biggest challenge with this project. This is not a simple problem and what we’re doing is dealing with it on a case by case basis. Some data sources are actually easier to disambiguate than others but at the moment we are using Zetoc which is a BL table of contents and that provides information like co-authors that have worked on a paper which is quite a good indicator of whether a person is the same person on another record within their database. In most cases we’re not saying that this person is 100% the person you are looking for but we’re giving an indication of how likely it is that this person is the same person. As part of this we are looking at how we can manage the records because there are going to be cases where we’ve got it wrong and one aspect of this might be author self-management – which has got its own problems. Allowing them to say that some of the information pertains to me but other parts are nothing to do with me and therefore split that off into a separate record. So, this is a really big problem, but that is the nature of the project.

Q: Bearing in mind that you’ll never guarantee 100% accuracy, how do you see this working in a repository, in that we recommend this authority name rather than something else blitzing and over-riding the information.
A: That is something we’re still trying to figure out at the moment. I think it probably will end up being a list of recommendations rather than ‘this is what you have to use’ because we don’t want to force that on anybody, but we do hope to give a pretty accurate list of potential matches.

Q: Would you agree that a lot of the problems will hopefully, in a way, be retrospective problems with the various attempts to address author identifiers? Going into the future I can see a situation where you’re actually in a cyber repository and the thing they’re actually holding is a persistent identifier relating to a person, with display names over the top to give a matter of preference as to how you display names.
A: I do agree. Although there is also an issue with the identifiers themselves, because if the record has an identifier that somebody claims – but the information pertains to two different people, then some of that information needs to be split off to a separate record. If two people are using the same identifier, there’s issues with the splitting and merging of records. It’s just another of the big issues that we’re trying to deal with at the moment.