WETA: Automatic taxonomy alignment via word embeddings

Giabelli, Anna; Malandri, Lorenzo; Mercorio, Fabio; Mezzanzanica, Mario

doi:10.1016/j.compind.2022.103626

Cited by 9 publications

(7 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Giabelli et al . (2022), the authors mentioned that:A taxonomy T is a 4-tuple T = ( C , W , H , F ).…”

Section: Resultsmentioning

confidence: 99%

“…In these 28 papers, we found at least four different variations of the definition of a taxonomy, quoted below. Maldonado [12] Merge-Into [13] Musgrove Pfeifer [15] Ponzetto [16] Reynaud [17] Taxo-Semantics [18] WETA [19] Xu [20] Application Areas Despite the many variations to define a taxonomy, there are many commonalities across all papers: taxonomy as a tree, taxonomy as a hierarchy, a taxonomy contains a set of concepts and a taxonomy contains a subset relation. Here, we reiterate our definition of a taxonomy in this study: A taxonomy T is defined as a pair: T 5 (C, E), where C is a set of concepts, and E is a set of edges.…”

Section: Definitions Of a Taxonomymentioning

confidence: 99%

“…Usually, the goal for employing an NLP approach is to enrich the source taxonomy or an existing dictionary by automating the taxonomy alignment processes. The taxonomies examined usually comprise a large number of concepts, and by automatically aligning two (or JD 79,6 more) taxonomies, the studies could come up more quickly with suggested candidate pairs that are ranked higher in similarity to be included in the updated taxonomy (Giabelli et al, 2022;Jung, 2006Jung, , 2008. Logic-based approaches.…”

Section: Taxonomy Alignment In Information Sciencesmentioning

confidence: 99%

“…):[15] Pfeifer and Peukert (2013a, b),[16] Ponzetto and Navigli (2009),[17] ,[18] Taxo-Semantics inAngermann et al (2017),[19] WETA inGiabelli et al (2022),[20] Source(s):…”

mentioning

confidence: 99%

“…represents a corpus of document. "InGiabelli et al (2022), the authors mentioned that:A taxonomy T is a 4-tuple T 5 (C, W, H, F). C is a set of concepts (aka, nodes) .…”

mentioning

confidence: 99%

See 4 more Smart Citations

A systematic review of methods for aligning, mapping, merging taxonomies in information sciences

Cheng

Xia

2023

View full text Add to dashboard Cite

PurposeThe purpose of this study is to provide a systematic literature review on taxonomy alignment methods in information science to explore the common research pipeline and characteristics.Design/methodology/approachThe authors implement a five-step systematic literature review process relating to taxonomy alignment. They take on a knowledge organization system (KOS) perspective, and specifically examining the level of KOS on “taxonomies.”FindingsThey synthesize the matching dimensions of 28 taxonomy alignment studies in terms of the taxonomy input, approach and output. In the input dimension, they develop three characteristics: tree shapes, variable names and symmetry; for approach: methodology, unit of matching, comparison type and relation type; for output: the number of merged solutions and whether original taxonomies are preserved in the solutions.Research limitations/implicationsThe main research implications of this study are threefold: (1) to enhance the understanding of the characteristics of a taxonomy alignment work; (2) to provide a novel categorization of taxonomy alignment approaches into natural language processing approach, logic-based approach and heuristic-based approach; (3) to provide a methodological guideline on the must-include characteristics for future taxonomy alignment research.Originality/valueThere is no existing comprehensive review on the alignment of “taxonomies”. Further, no other mapping survey research has discussed the comparison from a KOS perspective. Using a KOS lens is critical in understanding the broader picture of what other similar systems of organizations are, and enables us to define taxonomies more precisely.

show abstract

“…In Giabelli et al . (2022), the authors mentioned that:A taxonomy T is a 4-tuple T = ( C , W , H , F ).…”

Section: Resultsmentioning

confidence: 99%

Section: Definitions Of a Taxonomymentioning

confidence: 99%

Section: Taxonomy Alignment In Information Sciencesmentioning

confidence: 99%

“…):[15] Pfeifer and Peukert (2013a, b),[16] Ponzetto and Navigli (2009),[17] ,[18] Taxo-Semantics inAngermann et al (2017),[19] WETA inGiabelli et al (2022),[20] Source(s):…”

mentioning

confidence: 99%

“…represents a corpus of document. "InGiabelli et al (2022), the authors mentioned that:A taxonomy T is a 4-tuple T 5 (C, W, H, F). C is a set of concepts (aka, nodes) .…”

mentioning

confidence: 99%

See 3 more Smart Citations