Author name disambiguation of bibliometric data: A comparison of several unsupervised approaches

Adequately disambiguating author names in bibliometric databases is a precondition for conducting reliable analyses at the author level. In the case of bibliometric studies that include many researchers, it is not possible to disambiguate each single researcher manually. Several approaches have been proposed for author name disambiguation, but there has not yet been a comparison of them under controlled conditions. In this study, we compare a set of unsupervised disambiguation approaches. Unsupervised approaches specify a model to assess the similarity of author mentions a priori instead of training a model with labeled data. To evaluate the approaches, we applied them to a set of author mentions annotated with a ResearcherID, this being an author identifier maintained by the researchers themselves. Apart from comparing the overall performance, we take a more detailed look at the role of the parametrization of the approaches and analyze the dependence of the results on the complexity of the disambiguation task. Furthermore, we examine which effects the differences in the set of metadata considered by the different approaches have on the disambiguation results. In the context of this study, the approach proposed by Caron and van Eck (2014) produced the best results.

Greek chemical engineers. Are they static or mobile? Evidence from the national archive of PhD theses

Purpose The purpose of this paper is to identify the geographical location of researchers. Design/methodology/approach Combine standard bibliometric databases with social media data. Findings The majority of the population of the sample (71.8%) – Greek chemical engineers – are static. A significant portion of the mobile researchers (28.2%) returned to their country of origin (25.6%). Performing network analysis, the cluster of countries corresponding to the mobile category of researchers is identified and depicted. Originality/value Herein, this study introduce a new, national data set on doctorate holders that will allow multiple bibliometric analyses in the future. Also, this study is among the few (Gendronneau et al., 2019) that combines standard bibliometric databases with social media data. In cases where multiple affiliations per year pose a difficulty in understanding the geographical location of each individual, LinkedIn data were used. The analysis sheds light on a field of science that is not extensively examined in terms of brain circulation. While similar publications focus on physicians (i.e. cardiologists – Dyachenko and Mironenko, 2018), this paper focus on a subset of doctorate holders in engineering.

World Health Organization (WHO) COVID-19 Database: WHO Needs It?

A large number of COVID-19 publications has created a need to collect all research-related material in centralized databases. Generating and maintaining such databases regularly, while preserving the quality of the content is challenging, especially considering that the bibliometric databases rely on different data categorization strategies. In this short article, we investigate the functionality and quality of the WHO, PubMed and Scopus databases with a focus on missing values and duplicate entries related to COVID-19. Even though the WHO database is compiled from multiple sources, we conclude that using only the WHO database is not satisfactory as a lot of articles are still available exclusively in other databases. In addition to that, a more careful investigation revealed significant quality problems with all databases in terms of missing values, and many duplicate entries in the WHO database.

RESEARCH ON OPEN DIGITAL INFORMATION SYSTEMS FOR ANALYSIS OF RESEARCH ACTIVITY RESULTS OF HIGHER EDUCATION INSTITUTIONS ACADEMIC STAFF

In modern society there is a growing interest in assessing the quality of research activities of higher education institutions academic staff, in studying the possibilities of open digital information systems, the question regarding the indicators determination of scientific research effectiveness and their optimal representation in the global digital scientific space arises. The importance of analyzing the effectiveness of research results is due to the weight of publication activity in the most authoritative Ukrainian and international rankings. The article presents the results of the analysis of the Ukrainian and international experience in evaluating the research activities of higher education institution academic staff using indicators of international scientometric and bibliometric databases, analyzes the features of the most authoritative international scientometric and bibliometric databases - Scopus, Web of Science, Mendeley, Google Scholar, Research Gate, platforms for public access to research results – Zenodo, platforms for analyzing the prevalence of scientific publications on the Internet – Altmetric, their ability to search for publications and available tools for scientometric analysis of publishing activity of academic staff; the advantages and disadvantages of their use for the evaluation of research activity of higher education institution academic staff are identified. The peculiarities of the formation of research activity indicators of higher education institutions academic staff are determined, in particular from international scientometric and bibliometric databases, for optimal and complete analysis of their effectiveness.

Merits and Limits: Applying open data to monitor open access publications in bibliometric databases

Identifying and monitoring Open Access (OA) publications might seem a trivial task while practical efforts prove otherwise. Contradictory information arise often depending on metadata employed. We strive to assign OA status to publications in Web of Science (WOS) and Scopus while complementing it with different sourcesof OA information to resolve contradicting cases. We linked publications from WOS and Scopus via DOIs and ISSNs to Unpaywall, Crossref, DOAJ and ROAD. Only about 70% of articles and reviews from WOS and Scopus could be matched via a DOI to Unpaywall. Matching with Crossref brought 53 distinct licences, which define in many cases the legally binding access status of publications. But only 53% of publications hold only a single licence on Crossref, while more than 42% have no licence information submitted to Crossref. Contrasting OA information from Crossref licences with Unpaywall we found contradictory cases overall amounting to more than 13%, which might be partially explained by (ex-)including green OA. A further manual check found about 17% of OA publications that are not accessible and about 15% non-OA publications that are accessible through publishers’ websites. These preliminary results suggest that identification of OA state of publications denotes a difficult and currently unfulfilled task.

Study results for the detection of translated plagiarism using bibliometric databases

In the last five years the authors have studied applicability of citation analysis to the problem of detection of translated plagiarism. During these years we verified the efficiency of this method and carried out several applied studies. It enabled us to detect a number of illegal uses of borrowed foreign texts in several research areas. We showed that the method is more efficient in natural and exact sciences as compared to social sciences. Besides, it is more appropriate when analyzing research articles, reviews and reports as compared to other document types.