Persistent Identifiers - The 5 Things You Need To Know

Editor's Notes

• #6: We all have persistent identifiers, social security id, unique tax id, driving licence number, passport number and so forth. There is an abundance of identifiers for people, places and things, many have different purposes, some overlap, many have different metadata attached to them, but they should all hold three key attributes. A PID is usually numeric or alphanumeric. They should be: Unique, e.g. each entity has a single one of its type. If there are duplicate IDs for the same entity or party, how do you know which to use?Persistent, e.g. they do not change without very good cause, when a person changes name or their name is represented differently the ID should be the same – e.g. ORCID. When institutions merge or are acquired they should have the same ID and any subsumed parts should have their ID pointing to the absorbing organisation. Content IDs should be persistent and enable you to find the content over time, e.g. DOI. They should be meaningless so that if an attribute of the entity changes the ID does not need to, e.g. a person changing name, an organisation moving location. PIDs can be completely open and free to use, open under certain restrictions or proprietary. They must be attached to enough metadata to disambiguate one entity from another, but need no more than that. Many PIDs are attached to a large amount of metadata which support their use, this can include organisational hierarchies and classifications, authored research and contributions to research, publications, publishers, funders and much more. Things that do not count as PID include names, locations, affiliations, classifications and anything descriptive or that may change over time.
• #8: PIDs are used to disambiguate different entities that have the same name, acronym, title or version (e.g. VoR, Accepted Manuscript, Submitted Manuscript). The also avoid duplication, by using PIDs we can tell that the different spellings, naming convention or language are actually all the same entity. PIDs for organisations can also be organised into hierarchies linking the parts of an organisation together, which is particularly important when you wish to understand structure for entitlement, access and authentication, holdings, and so forth. Classifications attributed to organizations are also extremely useful and enable you to look at demographics, traits, gap analyses and much more.
• #9: Many organisations have started to use PIDs to link data together internally, mapping disparate systems and information sets together accurately for the first time. Beyond the boundaries of internal systems, PIDs are now widely used to accurately link data from one organisation to another throughout the supply chain. This can be about customers, e.g. subscribers to content, about author affiliations for bibliographic purposes or for the payment of APCs for open access journal articles. Furthermore, bridge identifiers can be used to pull proprietary datasets together – effectively to create mash-ups of information that had not been linked before. ISNI – is a bridge identifier which can enable such linkages as a bridge identifier it sits above proprietary and open PIDs and can enable them and the contents for their databases to be linked together.
• #11: It is really important that when working with PIDs that you are sure that they manage change effectively. Managing the changes in entities is time consuming but necessary, people change names, aliases and so forth, organisations merge, split, cease trading, move location, and content has lots of versions, we are probably going to start to see dynamic content emerging and managing the changes in that will be similar to other types of entities. So managing these changes must be Uniform, user of a PID system need to know that when something changes it changes in the same way. Maintenance of data connected to PIDs which gives them much of their utility needs to be Maintained. Things change all the time, organisation changes are probably much more frequent than you might expect. In order to ensure Uniformity of data and change management, it is important that PID datasets are Curated with some form of central management/corroboration
• #13: Using PIDs can save you money, understanding every relationship you have with an organisation saves time and effort in building relationships with your customers, and authors – it can also automate systems between organisations, including transactions and data flow. Data associated with PIDs can be used for many purposes in sales, marketing, customer services, editorial and helps pull knowledge together that is accessible by all where it used to sit in different departments. Research organizations are now able to link researchers together with common interests, locally, nationally and globally, help them locate relevant content, grants, even laboratories. Societies can look at their membership in relation to their authors, customers and editors. Ultimately unique, persistent and accurately maintained PIDs enable us all to trust data and utilise it to improve the relationships with the people and organisations on which we depend and which depend on us.
• #15: There are PIDs which apply to all kinds of entities – people, content, and places/organizations, and Ringgold IDs apply to institutions in the scholarly and research sphere. As laura mentioned, PIDs often have varying levels of data attached to them – an ISBN is linked to a book titlem for example. And our PID, the Ringgold ID, is linked toa robust set of descriptive data about that institution, and joined to other related Ringgold IDs, as I’ll show in a moment.
• #16: This Ringgold system of PIDs, data, & joined records is known as the Identify Database. Its primary purpose is to be a curated set of institutional records for our clients in the scholarly communications field, to aid in data governance and enable institutional data to be interoperable between internal systems, and with external partners like publishers, sales agents, intermediaries, and more. Because our clients may need master records for any kind of institutions, there are no geographic or sector boundaries on Identify. And while the database is pretty large now – we are closing in on half a million records – clients can work with us to have new records added, so that they have a complete set of PIDs and In addition to solving the problems of entity management and data interoperability with unique IDs, the supplemental metadata & hierarchies power business intelligence for our clients, which now number more thatn 80.
• #19: BRILL: Business Case: Create uniformity and reduce duplication of institutional data with Brill’s source data, in order to map data across multiple systems. Goals: • Create Intelligence – Analysis and Detailed reports • No future clean-up projects needed – Save employee’s time and costs • Record identity uniformity between Brill’s systems • Data control - Insight and control on the data (i.e paid subscriptions vs access delivered) • Better customer service due to Access controlled by unique Customer numbers (Ringgold ids) • Meaningful reports for records such as Authors or Individuals linked to a Unique Institutional ID. • Easier Data Migration and data management Primary systems included: Turpin – subscription management -Editorial Manager – peer review & manuscript submission system - also using Ringgold IDs -Klopotek – content management Still early days, but there are already positive changes as a result of embedding RINs in source data: • Tripled usage of Reporting tools, and eliminated outdated & clunky spreadsheets in favor of integrated reporting tools which did not enable knowledge to be easily shared among colleagues. • Decrease of Access claims - Less man power needed at our Customer Service dep. • Increased number of targeted contacts, with institutional affiliation, for our Marketing & Sales Department • Database & Report Accuracy: more than 70% of the records turned out to be duplicated • More rapid transition of customer data from one system to another.
• #20: Problem: Author affiliations An intuitive user interface enables users to select the correct institution from an Identify-populated search, embedding the Ringgold Identifier and Institution Name into the manuscript metadata or user profile. This feature is now active for all 6,400 journals served by ScholarOne Manuscripts, eliminating free text entries of author affiliations, and those of other contributors, reviewers, etc. Those Ringgold IDs are then transmitted in multiple ways: Housed in S1 systems holding profiles for authors, contributors reviewers. Sent to publishers in the Manuscript DTDs Publishers can also receive these IDs in reports, and can be then housed in their own systems – which is particularly powerful for the many publishers who also use the full Identify database CCC RLOA – which uses Ringgold’s accurate inst data to help calculate open access apcs.
• #21: Example of selection process for affiliation within S1M system – Ringgold ID is invisible to user, but used to enable definitive & accurate affiliation selection, and embed ID into users record/profile, and mss metadata.
• #22: There are many more examples of how PIDs & Ringgold Ids can help power business and connect your data, please visit us to learn more.