4D Specialty Approximation: Ability to Distinguish between Related Specialties

Nadine Rons
DOI: https://doi.org/10.48550/arXiv.1610.07496
2016-10-25
Abstract:Publication and citation patterns can vary significantly between related disciplines or more narrow specialties, even when sharing journals. Journal-based structures are therefore not accurate enough to approximate certain specialties, neither subject categories in global citation indices, nor cell sub-structures (Rons, 2012). This paper presents first test results of a new methodology that approximates the specialty of a highly specialized seed record by combining criteria for four publication metadata-fields, thereby broadly covering conceptual components defining disciplines and scholarly communication. To offer added value compared to journal-based structures, the methodology needs to generate sufficiently distinct results for seed directories in related specialties (sharing subject categories, cells, or even sources) with significantly different characteristics. This is tested successfully for the sub-domains of theoretical and experimental particle physics. In particular analyses of specialties with characteristics deviating from those of a broader discipline embedded in can benefit from an approach discerning down to specialty level. Such specialties are potentially present in all disciplines, for instance as cases of peripheral, emerging, frontier, or strategically prioritized research areas.
Digital Libraries
What problem does this paper attempt to address?