Author Name Disambiguation Using a New Categorical Distribution Similarity

Abstract

Abstract. Author name ambiguity has been a long-standing problem which impairs the accuracy of publication retrieval and bibliometric methods. Most of the existing disambiguation methods are built on similarity measures, e.g., "Jaccard Coefficient", between two sets of papers to be disambiguated, each set represented by a set of categorical features, e.g., coauthors and published venues 1 . Such measures perform bad when the two sets are small, which is typical in Author Name Disambiguation. In this paper, we propose a novel categorical set similarity measure. We model an author's preference, e.g., to venues, using a categorical distribution, and derive a likelihood ratio to estimate the likelihood that the two sets are drawn from the same distribution. This likelihood ratio is used as the similarity measure to decide whether two sets belong to the same author. This measure is mathematically principled and verified to perform well even when the cardinalities of the two compared sets are small. Additionally, we propose a new method to estimate the number of distinct authors for a given name based on the name statistics extracted from a digital library. Experiment shows that our method significantly outperforms a baseline method, a widely used benchmark method, and a real system.

A fast method based on multiple clustering for name disambiguation in bibliographic citations

“…The importance of the three attributes is decided by their impact on the name ambiguity problem. First, as a scholar usually has several related research fields and collaborates with several comparatively stable researchers in each field, coauthors usually provides stronger evidence than the other attributes (Cota, Ferreira, Nascimento, Gonçalves, & Laender, ; Fan, Wang, Pu, Zhou, & Lv, ; Kang et al., ; Li et al., ; Tang et al., ). We extract coauthors from article set and cluster the articles by coauthorship in the first step.…”

“…FMC consists of three steps: first, given the article list by authors with the same name, FMC groups articles into small clusters, also called fragments, by coauthorship where each cluster represents the articles of one author; second, FMC continues to cluster the fragments obtained from the previous step by correlating titles to reduce the number of fragments and increase the number of articles in the fragments; finally, the algorithm further tunes the clustering via the latent relations among venues, where the articles written by authors with the same name have been grouped together under the actual authors. Our experimental results show that FMC gets the best pairwise F1 score among four algorithms and reduces the runtime by 10 to 100 times as compared with the second‐best algorithm Categorical Sampling Likelihood Ratio (CSLR) (Li, Cong, & Miao, ).…”

Leveraging Co-authorship and Biographical Information for Author Ambiguity Resolution in DBLP

“…There are several researches conducted on mixed citation problem [4], and other ones deal with split citation problem [5]. Some approaches treat the both cases as in [15].…”

“…In [3] the authors proposes two attributes, the topic and the web correlation where measure the topic similarity across two citations and the use the web to detect whether two citations are listed together under the same web page which may flag a strong indicator that they belong to the same individual. In [4,5] the authors considers three features: co-authors, published venue and title where in categorical the authors investigate a categorical distribution similarity method to find the similarity among these features, and in k-way the authors used k-way spectral clustering for grouping the similar citations. In [6], the authors also used co-authors, title of publications in addition to the title of the proceedings or journal and the employ two supervised learning methods, SVM and naïve Bayes to classify two citations.…”

On Validation of Clustering Techniques for Bibliographic Databases

“…The problem is well known in the field of database and artificial intelligence. This problem is known with various other names like merge/purge problem [1], object distinction [2], author name ambiguity [3], temporal record linkage [4], [5], etc. Researchers have proposed various approaches to tackle this problem.…”