A knowledge graph embeddings based approach for author name disambiguation using literals

• DOI: 10.1007/s11192-022-04426-2
• Publication Date: 2022-07-04T19:02:32Z
• Authors: Cristian Santini; Genet Asefa Gesese; Silvio Peroni; Aldo Gangemi; Harald Sack; Mehwish Alam
• Keywords: Benchmark (surveying); Knowledge graph; Literal (mathematical logic); Code (set theory); Hierarchical clustering

Abstract Scholarly data is growing continuously containing information about the articles from a plethora of venues including conferences, journals, etc. Many initiatives have been taken to make scholarly available in form Knowledge Graphs (KGs). These efforts standardize these and them accessible also led many challenges such as exploration articles, ambiguous authors, This study more specifically targets problem Author Name Disambiguation (AND) on KGs presents novel framework, Literally (LAND), which utilizes Graph Embeddings (KGEs) using multimodal literal generated KGs. framework based three components: (1) KGEs, (2) blocking procedure, finally, (3) hierarchical Agglomerative Clustering. Extensive experiments conducted two newly created KGs: (i) KG Scientometrics Journal 1978 onwards (OC-782K), (ii) extracted well-known benchmark for AND provided by AMiner (AMiner-534K). The results show that our proposed architecture outperforms baselines 8–14% terms F 1 score shows competitive performances challenging AMiner. code datasets are publicly through Github ( https://github.com/sntcristian/and-kge ) Zenodo https://doi.org/10.5281/zenodo.6309855 respectively.