<scp>Ethnicity‐based</scp> name partitioning for author name disambiguation using supervised machine learning

Kim, Jinseok; Kim, Jenna; Owen‐Smith, Jason

doi:10.1002/asi.24459

Cited by 10 publications

(4 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, the consideration of individual publication histories at the level of authors or gender analysis requires name disambiguation. Obviously, this task is a big challenge, especially when dealing with non-Western names (Gomide et al, 2017;Kim et al, 2021;Treeratpituk & Giles, 2012). Since the majority of publications in the Ukrainian Economic discipline are related to local authors (Mryglod et al, 2021) it is natural to find mainly Ukrainian first and last names in our data set.…”

Section: Research Papermentioning

confidence: 99%

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①

Mryglod

Nazarovets

Kozmenko

2023

Journal of Data and Information Science

View full text Add to dashboard Cite

Purpose To supplement the quantitative portrait of Ukrainian Economics discipline with the results of gender and author ordering analysis at the level of individual authors, special methods of working with bibliographic data with a predominant share of non-English authors are used. The properties of gender mixing, the likelihood of male and female authors occupying the first position in the authorship list, as well as the arrangements of names are studied. Design/methodology/approach A data set containing bibliographic records related to Ukrainian journal publications in the field of Economics is constructed using Crossref meta-data. Partial semi-automatic disambiguation of authors’ names is performed. First names, along with gender-specific ethnic surnames, are used for gender disambiguation required for further comparative gender analysis. Random reshuffling of data is used to determine the impact of gender correlations. To assess the level of alphabetization for our data set, both Latin and Cyrillic versions of names are taken into account. Findings The lack of well-structured metadata and the poor use of digital identifiers lead to numerous problems with automatization of bibliographic data pre-processing, especially in the case of publications by non-Western authors. The described stages for working with such specific data help to work at the level of authors and analyse, in particular, gender issues. Despite the larger number of female authors, gender equality is more likely to be reported at the individual level for the discipline of Ukrainian Economics. The tendencies towards collaborative or solo-publications and gender mixing patterns are found to be dependent on the journal: the differences for publications indexed in Scopus and/or Web of Science databases are found. It has also been found that Ukrainian Economics research is characterized by rather a non-alphabetical order of authors. Research limitations Only partial authors’ name disambiguation is performed in a semi-automatic way. Gender labels can be derived only for authors declared by full First names or gender-specific Last names. Practical implications The typical features of Ukrainian Economic discipline can be used to perform a comparison with other countries and disciplines, to develop an informed-based assessment procedure at the national level. The proposed way of processing publication data can be borrowed to enrich metadata about other research disciplines, especially for non-English speaking countries. Originality/value To our knowledge, this is the first large-scale quantitative study of Ukrainian Economic discipline. The results obtained are valuable not only at the national level, but also contribute to general knowledge about Economic research, gender issues, and authors’ names ordering. An example of the use of Crossref data is provided, while this data source is still less used due to a number of drawbacks. Here, for the first time, attention is drawn to the explicit use of the features of the Slavic authors’ names.

show abstract

Section: Research Papermentioning

confidence: 99%

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①

Mryglod

Nazarovets

Kozmenko

2023

Journal of Data and Information Science

View full text Add to dashboard Cite

show abstract

“…The work on name separation using machine learning is much more limited than the work for ethnicity classification using machine learning. Kim et al [25] tackled the problem of distinguishing authors who have the same names (i.e., same name means same first forename initial and full surname) in bibliographic. They used ethnicity information for better disambiguation performance.…”

Section: Ethnicity Classification and Name Separationmentioning

confidence: 99%

Malaysian Name-based Ethnicity Classification using LSTM

2022

KSII TIIS

View full text Add to dashboard Cite

Name separation (splitting full names into surnames and given names) is not a tedious task in a multiethnic country because the procedure for splitting surnames and given names is ethnicity-specific. Malaysia has multiple main ethnic groups; therefore, separating Malaysian full names into surnames and given names proves a challenge. In this study, we develop a twophase framework for Malaysian name separation using deep learning. In the initial phase, we predict the ethnicity of full names. We propose a recurrent neural network with long shortterm memory network-based model with character embeddings for prediction. Based on the predicted ethnicity, we use a rule-based algorithm for splitting full names into surnames and given names in the second phase. We evaluate the performance of the proposed model against various machine learning models and demonstrate that it outperforms them by an average of 9%. Moreover, transfer learning and fine-tuning of the proposed model with an additional dataset results in an improvement of up to 7% on average.

show abstract

“…Specifically, in WhoisWho, most last names (53 out of 65) are Chinese last names, which is inconsistent with the fact that the authors are from all over the world. Note that this issue is nontrivial because many studies have confirmed that different ethnicities have different levels of ambiguities (Louppe et al, 2016), and, based on this idea, some ethnicity‐based disambiguation methods have been successfully developed (Kim et al, 2021; Louppe et al, 2016; Subramanian et al, 2021). Name variation is another frequently ignored aspect in building AND datasets.…”

Section: Related Workmentioning

confidence: 99%

LAGOS‐AND: A large gold standard dataset for scholarly author name disambiguation

Zhang

Yang

2022

Asso for Info Science & Tech

View full text Add to dashboard Cite

In this article, we present a method to automatically build large labeled datasets for the author ambiguity problem in the academic world by leveraging the authoritative academic resources, ORCID and DOI. Using the method, we built LAGOS‐AND, two large, gold‐standard sub‐datasets for author name disambiguation (AND), of which LAGOS‐AND‐BLOCK is created for clustering‐based AND research and LAGOS‐AND‐PAIRWISE is created for classification‐based AND research. Our LAGOS‐AND datasets are substantially different from the existing ones. The initial versions of the datasets (v1.0, released in February 2021) include 7.5 M citations authored by 798 K unique authors (LAGOS‐AND‐BLOCK) and close to 1 M instances (LAGOS‐AND‐PAIRWISE). And both datasets show close similarities to the whole Microsoft Academic Graph (MAG) across validations of six facets. In building the datasets, we reveal the variation degrees of last names in three literature databases, PubMed, MAG, and Semantic Scholar, by comparing author names hosted to the authors' official last names shown on the ORCID pages. Furthermore, we evaluate several baseline disambiguation methods as well as the MAG's author IDs system on our datasets, and the evaluation helps identify several interesting findings. We hope the datasets and findings will bring new insights for future studies. The code and datasets are publicly available.

show abstract

Ethnicity‐based name partitioning for author name disambiguation using supervised machine learning

Cited by 10 publications

References 45 publications

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①

Malaysian Name-based Ethnicity Classification using LSTM

LAGOS‐AND: A large gold standard dataset for scholarly author name disambiguation

Contact Info

Product

Resources

About

Ethnicity‐based name partitioning for author name disambiguation using supervised machine learning

Cited by 10 publications

References 45 publications

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases①

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases①

Malaysian Name-based Ethnicity Classification using LSTM

LAGOS‐AND: A large gold standard dataset for scholarly author name disambiguation

Contact Info

Product

Resources

About

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①

Peculiarities of gender disambiguation and ordering of non-English authors’ names for Economic papers beyond core databases^①