Search (4 results, page 1 of 1)

0.0028703054 = product of:
  0.005740611 = sum of:
    0.005740611 = product of:
      0.011481222 = sum of:
        0.011481222 = weight(_text_:a in 6345) [ClassicSimilarity], result of:
          0.011481222 = score(doc=6345,freq=16.0), product of:
            0.053105544 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046056706 = queryNorm
            0.2161963 = fieldWeight in 6345, product of:
              4.0 = tf(freq=16.0), with freq of:
                16.0 = termFreq=16.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046875 = fieldNorm(doc=6345)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

Five years ago, we raised the question: is there a data model which is general enough such that all existing thesauri can be represented by a specialization of this general model without loss of information? The answer was not given at that time, but we referred to the principle of abstraction, well supported in object-oriented data modelling. We gained the empirical basis for that process of abstraction by modelling existing thesauri and a terminological dictionary; an abstracting view was afterwards presented in a paper to the TKE'93 conference. The present paper reports on a third step of abstraction with its very concrete consequences, embodies in a software called TerminologyFramework(TFw)

Type

a

0.0026473717 = product of:
  0.0052947435 = sum of:
    0.0052947435 = product of:
      0.010589487 = sum of:
        0.010589487 = weight(_text_:a in 2362) [ClassicSimilarity], result of:
          0.010589487 = score(doc=2362,freq=40.0), product of:
            0.053105544 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046056706 = queryNorm
            0.19940455 = fieldWeight in 2362, product of:
              6.3245554 = tf(freq=40.0), with freq of:
                40.0 = termFreq=40.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.02734375 = fieldNorm(doc=2362)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

The paper analysed here is a kind of position paper. In order to get a better under-standing of the reported work I used the retrieval interface of the thesaurus, the so-called NCI DTS Browser accessible via the Web3, and I perused the cited OWL file4 with numerous "Find" and "Find next" string searches. In addition the file was im-ported into Protégé 2000, Release 2.0, with OWL Plugin 1.0 and Racer Plugin 1.7.14. At the end of the paper's introduction the authors say: "In the following sections, this paper will describe the terminology development process at NCI, and the issues associated with converting a description logic based nomenclature to a semantically rich OWL ontology." While I will not deal with the first part, i.e. the terminology development process at NCI, I do not see the thesaurus as a description logic based nomenclature, or its cur-rent state and conversion already result in a "rich" OWL ontology. What does "rich" mean here? According to my view there is a great quantity of concepts and links but a very poor description logic structure which enables inferences. And what does the fol-lowing really mean, which is said a few lines previously: "Although editors have defined a number of named ontologic relations to support the description-logic based structure of the Thesaurus, additional relation-ships are considered for inclusion as required to support dependent applications."
According to my findings several relations available in the thesaurus query interface as "roles", are not used, i.e. there are not yet any assertions with them. And those which are used do not contribute to complete concept definitions of concepts which represent thesaurus main entries. In other words: The authors claim to already have a "description logic based nomenclature", where there is not yet one which deserves that title by being much more than a thesaurus with strict subsumption and additional inheritable semantic links. In the last section of the paper the authors say: "The most time consuming process in this conversion was making a careful analysis of the Thesaurus to understand the best way to translate it into OWL." "For other conversions, these same types of distinctions and decisions must be made. The expressive power of a proprietary encoding can vary widely from that in OWL or RDF. Understanding the original semantics and engineering a solution that most closely duplicates it is critical for creating a useful and accu-rate ontology." My question is: What decisions were made and are they exemplary, can they be rec-ommended as "the best way"? I raise strong doubts with respect to that, and I miss more profound discussions of the issues at stake. The following notes are dedicated to a critical description and assessment of the results of that conversion activity. They are written in a tutorial style more or less addressing students, but myself being a learner especially in the field of medical knowledge representation I do not speak "ex cathedra".

0.0024857575 = product of:
  0.004971515 = sum of:
    0.004971515 = product of:
      0.00994303 = sum of:
        0.00994303 = weight(_text_:a in 2176) [ClassicSimilarity], result of:
          0.00994303 = score(doc=2176,freq=12.0), product of:
            0.053105544 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046056706 = queryNorm
            0.18723148 = fieldWeight in 2176, product of:
              3.4641016 = tf(freq=12.0), with freq of:
                12.0 = termFreq=12.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046875 = fieldNorm(doc=2176)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Abstract

The ISO 2788 guidelines for monolingual thesauri contain a differentiation of "the hierarchical relationship" into "generic", "partitive", and "instance", which, for purposes of document retrieval, was deemed adequate. However, ontologies, designed as language inventories for a wider scope of knowledge representation, are based on all these and some more logical differentiations. Rereading the ISO 2788 standard and inspecting the published Cyc Upper Ontology, it is argued that the adoption of the document-retrieval definition of subsumption generally prevents the conception or use of a thesaurus as a substructure of an ontology of the new kind as constructed for AI applications. When a thesaurus is used for fact description and inference on fact descriptions, the instance-of relationship too should be reconsidered: It may also link concepts and metaconcepts, and then its distinction from subsumption is needed. The treatment of the instance-of relationship in thesauri, the Cyc Upper Ontology, and WordNet is described from this perspective

Type

a

0.0023678814 = product of:
  0.0047357627 = sum of:
    0.0047357627 = product of:
      0.009471525 = sum of:
        0.009471525 = weight(_text_:a in 23) [ClassicSimilarity], result of:
          0.009471525 = score(doc=23,freq=2.0), product of:
            0.053105544 = queryWeight, product of:
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.046056706 = queryNorm
            0.17835285 = fieldWeight in 23, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              1.153047 = idf(docFreq=37942, maxDocs=44218)
              0.109375 = fieldNorm(doc=23)
      0.5 = coord(1/2)
  0.5 = coord(1/2)

Type

a