What are we doing?

Starting August 2026, new records across INSPIRE will use only the parent-form of the institution name, e.g., “Harvard U. (main)” instead of “Harvard U., Dept. Math.” or “Cambridge U. (main)” instead “Cambridge U., DAMTP”. Eventually, once the clean-up has been completed, the “(main)” will be dropped and the final form of the name will be, e.g., “Harvard U.”.

Why are we doing this?

The somewhat inconsistent granularity of INSPIRE institution records dates back to the SPIRES days, when SLAC would send announcements of each week’s preprints through postal mail to institutions around the world. To do this, the mail needed to be addressed at the department level, and so the Physics department was the default for each institution’s record. To cover scientists who were in other departments, additional institution records would be created for, e.g., Mathematics departments.

Over time records proliferated in an unsustainable way as more people wanted their particular department added and as departments changed names and universities added centers, institutes and other sub-units. This often led to precision without accuracy as catalogers would misidentify the sub-unit and authors themselves would not be consistent about what they used for their affiliation.

In the era of automation, this problem became increasingly acute as the more choices there were, the greater possibility for errors that would then propagate across hundreds of records. This also produces complications for author identification as the same person may appear, meaninglessly, at different sub-units.

Having only one record for each institution simplifies things and allows INSPIRE to concentrate on producing accurate records and author identification.

TIB – Leibniz Information Centre for Science and Technology and University Library of Hannover has joined the INSPIRE collaboration and will be taking on some of the responsibilities previously held by DESY.

In particular, TIB will work on content selection, i.e., deciding which articles ought to be included in the INSPIRE-HEP literature database. Over a two-year period, TIB seeks to increase automation in INSPIRE’s content selection workflows, leveraging state-of-the-art machine learning algorithms. In doing so, TIB places a strong focus on quality assurance, keeping human experts in the loop for quality-sensitive tasks. In addition to content selection, TIB also contributes to the curation of INSPIRE-HEP metadata (adding information on author affiliations, conferences, experiments, etc.)

About TIB

TIB provides scientists, researchers, industry and business with literature and information. TIB holds a unique science and technology collection, including non-traditional knowledge objects such as audiovisual media and research data.

In addition to its roles as Germany’s national library of science and technology and its local services as the University Library of Leibniz Universität Hannover, TIB plays an active role in the digitalization of science and technology:

· As an information centre, TIB develops robust digital infrastructures and services and conducts its own research.

· As a public institution, TIB is committed to openness in science.

· As a research library, TIB’s mission is to facilitate access to scientific information and to strengthen the exchange of knowledge through applied research and development.