Big Scholarly Data im Open Access Monitor: ein Werkstattbericht

Stanzel, Franziska; Pollack, Philipp; Barbers, Irene; Lindstrot, Barbara

doi:10.18452/24797

Journal Article

FZJ-2022-02522

Big Scholarly Data im Open Access Monitor: ein Werkstattbericht

Stanzel, F. (Corresponding author)FZJ* ; Barbers, I.FZJ* ; Pollack, P.FZJ* ; Lindstrot, B.FZJ*

2022
Institut für Bibliotheks- und Informationswissenschaft der Humboldt Universität zu Berlin Berlin

LIBREAS. Library ideas 41, 20 pages (2022) [10.18452/24797]

This record in other databases:

Please use a persistent id in citations: http://hdl.handle.net/2128/31401 doi:10.18452/24797

Abstract: In the light of the Open Access transformation, the analysis of large amounts of data is increasingly important for libraries, whereas the number of scholarly publications is constantly growing. Large amounts of data must first be made usable before any substantiated analysis can be made, e.g. regarding institution-related publication outputs. This is where the Open Access Monitor (OAM) comes in, which acts as an interface for merging data from various source systems such as Unpaywall, Dimensions, Web of Science and Scopus. For this purpose, the OAM is structurally divided into three parts: the backend hosts the data, which can be queried via the API, and is presented and visualized in the frontend. All data, coming from various source systems, must be homogenized in order to realize complete data sets without creating duplicates. Journal titles or institution names have to be standardized to allow assigning the original entries from the source systems to the corresponding data records in the OAM. In the case of institution names, these are enriched with persistent identifiers. Given the way the data is organized in some of the source databases, the institution names cannot be mapped directly to organization identifiers (ROR-IDs) in some cases. Therefore, the raw forms of the author’s affiliation information are used in the mapping process. Affiliation mapping is an extensive and complex task, since the data provided are often ambiguous and at the same time a clear distinction of institutions, especially in the case of university hospitals, requires intellectual processing. The highly complex process of generating a uniform data set from a multitude of data sources will be demonstrated, with a special focus on the normalization processes as well as the assignment of Open Access categories. Metadata quality remains a constant challenge, as does the issue of availability and sustainability of the connected source systems. The use and integration of open data sources is generally desirable – it would be in line with the OAM’s goal of unrestricted (re-) usability of the OAM data. The pros and cons of using non-commercial databases are discussed using OpenAlex as an example.

Keyword(s): Publikationsdaten ; Anwendung ; wissenschaftliches Publizieren ; Open Access ; Monitoring ; open access monitoring ; publication data ; application ; scholarly publishing ; scholarly communication