Document Type
Article
Publication Title
Semantic Web: – Interoperability, Usability, Applicability
Abstract
We present a novel, logic-based solution to the challenge of reconciling the meanings of taxonomic names across multiple biological taxonomies. The challenge arises due to limitations inherent in using type-anchored taxonomic names as identifiers of granular semantic similarities and differences being expressed in original and revised taxonomic classifications. We address this challenge through: (1) the use of taxonomic concept labels – thereby individuating name usages according to particular sources and allowing each taxonomy to be recognized separately; (2) sets of user-provided Region Connection Calculus articulations among concepts (RCC-5: congruence, proper inclusion, inverse proper inclusion, overlap, exclusion); and (3) the use of an Answer Set Programming-based reasoning toolkit that ingests these constraints to infer and visualize consistent multi-taxonomy alignments. The feasibility of this approach is demonstrated with a use case involving pairwise alignments of 11 non-congruent classifications of Eastern United States grass entities variously assigned to the Andropogon glomeratus-virginicus ‘complex’ over an interval of 126 years. Analyses of name:meaning identity reveal that, on average, taxonomic names are reliable identifiers of taxonomic non-/congruence for approximately 60% of the 127 merge regions obtained in 12 pairwise alignments. The name:meaning cardinality over the entire time interval ranges from 1:6 to 4:1, with only 1:36 names attaining the semantically ideal 1:1 ratio. We discuss the applicability of the RCC-5 alignment approach in the context of achieving logic-based integration of non-/congruent taxonomic concept hierarchies in dynamic biodiversity data environments.
Pages
645-667
html
DOI
https://doi.org/10.3233/SW-160220
Volume
7
Issue
6
Publication Date
2016
Keywords
Answer set programming, biodiversity data, biological nomenclature, knowledge representation, ontology alignment
Disciplines
Computer Sciences | Data Science
ISSN
1570-0844
Recommended Citation
Franz, N. M., Chen, M., Kianmajd, P., Yu, S., Bowers, S., Weakley, A. S., & Ludäscher, B.. (2016). Names are not good enough: Reasoning over taxonomic change in the Andropogon complex. Semantic web, 7(6), 645–667. https://doi.org/10.3233/SW-160220
Upload File
wf_yes
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License