Search (1 results, page 1 of 1)

0.010770457 = product of:
  0.043081827 = sum of:
    0.043081827 = weight(_text_:digital in 1377) [ClassicSimilarity], result of:
      0.043081827 = score(doc=1377,freq=2.0), product of:
        0.19770671 = queryWeight, product of:
          3.944552 = idf(docFreq=2326, maxDocs=44218)
          0.050121464 = queryNorm
        0.21790776 = fieldWeight in 1377, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.944552 = idf(docFreq=2326, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1377)
  0.25 = coord(1/4)

Abstract

The organization of large collections of documents has become more important with the increase in the amount of digital information available. In certain constricted domains of knowledge, keywords and subject descriptors tend to be similar and therefore insufficient to differentiate documents. In this context, instead of relying only on the presence of common terms, the identification of common cited references can be useful to define semantic relationship among documents. The purpose of this work is to add another instance on the research linking information retrieval and bibliometric techniques aided by information technology. A domain analytic method was developed to generate clusters of documents, which uses self-organizing maps, in the scope of artificial neural networks, to categorize documents. The results obtained show that this approach successfully identified clusters of authors and documents through their cited references. In addition, further qualitative analysis of these clusters demonstrates the existence of semantic relationships between the documents. This study can contribute to the development of the field of knowledge organization by evaluating the use of artificial neural networks in the automatic categorization of documents in a constricted knowledge domain based on the analysis of the references cited by these documents.