New Physics Classification Scheme Unveiled

By Emily Conover

Watch out physicists: There’s a new taxonomy in town. At the beginning of January 2016, APS unveiled the Physics Subject Headings (PhySH) classification system, which will replace the Physics and Astronomy Classification Scheme (PACS) previously used to organize APS journal articles into subject areas. The new system will serve as a tool for readers of the APS journals — allowing for easier navigation through papers and topics of interest — and will help APS assign papers to editors.

The American Institute of Physics (AIP) — an organization of scientific member societies, of which APS is one — created PACS in the 1970s. However, in 2010, AIP decided to stop maintaining it. A statement on AIP Publishing’s website says, “The continuing evolution of indexing, search, and technology has brought into focus the inherent limitations of PACS and as a result, AIP decided that PACS 2010 would be the final version.”

This left a void, with no up-to-date classification system available for APS journals. “Of course new concepts in physics are appearing all the time, so it was starting to be problematic for new fields of physics,” says APS Chief Information Officer Mark Doyle. Topological insulators, for example, are not included in PACS. As a result, Doyle says, “We decided we were going to build our own.” The system has been under development since 2012.

PhySH categorizes papers according to “concepts,” which are tags that indicate the subject matter of the paper and the discipline it belongs to. During submission, authors select these from a list of about 3000. These concepts belong to “facets” — broad categories that indicate the nature of the concept. The facets currently include research areas, physical systems, properties, techniques (computational, experimental, and theoretical), and professional topics. “The idea is to try to distinguish the ‘what’ and the ‘how’ from each other,” Doyle says. Concepts also belong to disciplines — subfields of study within physics, such as nuclear physics, biological physics, or fluid dynamics. These two dimensions by which concepts are organized (facets and disciplines) allow readers to browse articles across subject matter, techniques, disciplines, or other qualities.

Concepts can belong to more than one discipline and more than one facet. Solitons, for example, belong to two facets: research areas and physical systems, as well as a number of disciplines: condensed matter and materials physics; fluid dynamics; nonlinear dynamics; particles and fields; plasma physics; and statistical physics.

The concepts also follow a hierarchy going from broader to narrower subject matter. Loop quantum gravity is a narrower concept than quantum gravity, for example. Concepts are also connected to other, related concepts. For example, spontaneous symmetry breaking is associated with the related concepts of symmetries and Higgs bosons.

The faceted hierarchy system should make it easier to browse papers and understand the connections between related papers and topics, Doyle says, as compared to PACS, which has a more rigid hierarchy.

An additional challenge with PACS, Doyle says, is that it was not designed for the modern web environment. PhySH is designed to support “linked data,” which would allow computers to read and analyze data about the papers.

PhySH is also designed to be simpler for users to understand. “The huge plus is that it’s word-based; there are actual English words rather than cryptic alphanumeric codes,” says Abhishek Agarwal, associate editor for Physical Review Letters, who helped coordinate the development of PhySH. “It’s much more intuitive; it’s transparent.”

Currently, APS is still collecting PACS with submitted articles because internal peer review processes depend on them, but authors are now able to add PhySH terms voluntarily. (For the newest APS journal, Physical Review Fluids, authors are required to select PhySH terms.) In the coming months, APS journals will phase out the use of PACS entirely.

Agarwal emphasizes that PhySH should not yet be considered a finished project, but, he says the feedback so far is “extremely encouraging.”

“The information that’s coming in from the authors is very accurate so far,” Agarwal says. “It’s very early days, but authors are using it and they are using it largely the way we envisioned they would use it.”

APS editors plan to continue refining the system with input from the physics community. “It’s still very much under development,” says Doyle. “We’re looking for feedback on it. We’re looking to iterate.”

More information about PhySH is available at the PhySH website.