Semantic network

A semantic network is a directed graph consisting of nodes (also termed points or vertices) which represent concepts and edges (also termed lines or arcs) which represent semantic relations between the concepts. A kind of knowledge representation used, for example, in hypertext systems.


 

A semantic network is a way of representing relationships between concepts. Often each concept is represented by a word or set of words. A simple example is a hierarchical network where the concepts are taxonomic terms from biology, and the only type of relationship is type-of (hyponymous relationship). More complex semantic networks many include a variety of types of relationship such as hardness, temperature, made-of, texture and color. One of the largest existing semantic networks is WordNet, a lexical database for the English language.

The difference between semantic networks and semantic tools such as thesauri developed in Library and Information Science (LIS) is discussed by Gilchrist:

"In parallel with advances in thesaurus compilation and manipulation, work was being undertaken by workers in artificial Intelligence in their construction of expert systems. In this work they compiled what are called 'semantic networks' described by Milstead (1995) as being "conceptually quite similar to  a thesaurus, in that they show terms in the context of their semantic relationships. However, they offer different navigation capabilities, through graphic devices that represent multidimensional spaces, rather than through review of the cross references and scope notes of a thesaurus". . . . another major difference between them [semantic networks] and thesauri was in the fact that the former went much further in defining the types of relationships between terms. For example, Ford (1991) in a textbook specially written for librarians and information scientists, presents a section of a semantic map devoted to cardiovascular illness, in which "heart disease" is shown as being "is-a" "cardiovascular illness" (is-a being a standard convention), which "has setting" "cardiovascular system" while "heart disease" 'has setting' 'heart', which is 'component of' 'cardiovascular system'. Two things arise from this more complex treatment. The first is that there is no intrinsic reason why a conventional thesaurus should not be extended and elaborated to include, for example, term definitions, notes on term usage, and more explicitly defined relationships. The second is that such enrichment allows the semantic network to be more easily manipulated by an inference engine, typically employing the IF . . . THEN operator. This enrichment of thesauri (see, e. g. Hazewinkel, 1997) is a feature of work in the area of ontologies, as will be shown in a later section. " (Gilchrist, 2003, p. 9-10).

The application within LIS seems so far to be limited to the UMLS project of medical terminology. See: Burgun & Bodenreider (2001), Halper et al. (2001), Yu et al. (1999), Zhang et al. (2004, 2005).

Literature:

Burgun, A. & Bodenreider, O. (2001). Mapping the UMLS semantic network into general ontologies. Journal of the American Medical Informatics Association, 2001,S, 81-85.

Cohen, P. R. & Kjeldsen, R. (1987). Information retrieval by constrained spreading activation in semantic networks. Information Processing & Management, 23(4), 255-268.

Ford, N . (1991). Expert Systems and Artificial Intelligence. London: Library Association Publishing.
 

Gilchrist, A (2003). Thesauri, taxonomies and ontologies - an etymological note. Journal of Documentation 59(1), 7-18.

Halper M ; Chen Z; Geller J; Perl, Y. (2001). A metaschema of the UMLS based on a partition of its semantic network. Journal of the American Medical Informatics Association, S, 234-238.

Hazewinkel, M. (1996/1997). Enriched thesauri and their uses in information retrieval and storage. Discussion paper. IN: Thanos, C .( Ed.). Proceedings of the First DELOS
Workshop, March 1996, ERCIM, pp. 27-32. (Revised 1997). Available at: http://www.ercim.org/publication/ws-proceedings/DELOS1/hazewinkel.pdf

Milstead, J. (1995). Invisible thesauri: the year 2000. Online & CDROM Review, 19(2),
93-94.

Rada, R. (1990). Hypertext writing and document reuse: the role of a semantic net. Electronic Publishing Review, 3(3), 125-140.
 

Ritchie, G. D. & Hanna, F. K. (1983). Semantic networks - a general definition and a survey. Information Technology: Research and Development, 2(4), 187-231.

Sowa, J. F. (2006). Semantic networks. (Revised and extended version of an article originally written for the Encyclopedia of Artificial Intelligence, edited by Stuart C. Shapiro, Wiley, 1987, second edition, 1992). http://www.jfsowa.com/pubs/semnet.htm

Wikipedia, the free encyclopedia.(2005). Semantic network. http://en.wikipedia.org/wiki/Semantic_network

Wikipedia, the free encyclopedia.(2006). Graph (mathematics). http://en.wikipedia.org/wiki/Graph_%28mathematics%29

Yu, H.; Friedman C; Rhzetsky A; Kra, P. (1999). Representing genomic knowledge in the UMLS semantic network. Journal of the American Medical Informatics Association, S, 181-185.

Zhang L ; Perl Y; Halper M; Geller J; Cimino, J. J. (2004). An enriched Unified Medical Language System Semantic Network with a multiple subsumption hierarchy. Journal of the American Medical Informatics Association, 11(3), 195-206.

Zhang L; Halper M; Perl Y ; Geller J; Cimino, J. J. (2005). Relationship structures and semantic type assignments of the UMLS Enriched Semantic Network. Journal of the American Medical Informatics Association, 12(6), 657-666.

----

WordNet: http://www.cogsci.princeton.edu/~wn/

Semantic relations in WordNet: http://www.cs.ucl.ac.uk/staff/a.hunter/tradepress/wordnet.html

http://www.informatics.susx.ac.uk/books/computers-and-thought/chap6/node5.html

Birger Hjørland

Last edited: 16-07-2006

Home