Search (2 results, page 1 of 1)

0.0528507 = product of:
  0.1057014 = sum of:
    0.1057014 = sum of:
      0.07819214 = weight(_text_:tagging in 3644) [ClassicSimilarity], result of:
        0.07819214 = score(doc=3644,freq=2.0), product of:
          0.342494 = queryWeight, product of:
            5.9038734 = idf(docFreq=327, maxDocs=44218)
            0.058011748 = queryNorm
          0.2283022 = fieldWeight in 3644, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            5.9038734 = idf(docFreq=327, maxDocs=44218)
            0.02734375 = fieldNorm(doc=3644)
      0.027509268 = weight(_text_:22 in 3644) [ClassicSimilarity], result of:
        0.027509268 = score(doc=3644,freq=2.0), product of:
          0.20314726 = queryWeight, product of:
            3.5018296 = idf(docFreq=3622, maxDocs=44218)
            0.058011748 = queryNorm
          0.1354154 = fieldWeight in 3644, product of:
            1.4142135 = tf(freq=2.0), with freq of:
              2.0 = termFreq=2.0
            3.5018296 = idf(docFreq=3622, maxDocs=44218)
            0.02734375 = fieldNorm(doc=3644)
  0.5 = coord(1/2)

Abstract

Calvin Mooers' work toward the resolution of the problem of ambiguity in indexing went unrecognized for years. At the time he introduced the "descriptor" - a term with a very distinct meaning-indexers were, for the most part, taking index terms directly from the document, without either rationalizing them with context or normalizing them with some kind of classification. It is ironic that Mooers' term came to be attached to the popular but unsophisticated indexing methods which he was trying to root out. Simply expressed, what Mooers did was to take the dictionary definitions of terms and redefine them so clearly that they could not be used in any context except that provided by the new definition. He did, at great pains, construct such meanings for over four hundred words; disambiguation and specificity were sought after and found for these words. He proposed that all indexers adopt this method so that when the index supplied a term, it also supplied the exact meaning for that term as used in the indexed document. The same term used differently in another document would be defined differently and possibly renamed to avoid ambiguity. The disambiguation was achieved by using unabridged dictionaries and other sources of defining terminology. In practice, this tends to produce circularity in definition, that is, word A refers to word B which refers to word C which refers to word A. It was necessary, therefore, to break this chain by creating a new, definitive meaning for each word. Eventually, means such as those used by Austin (q.v.) for PRECIS achieved the same purpose, but by much more complex means than just creating a unique definition of each term. Mooers, however, was probably the first to realize how confusing undefined terminology could be. Early automatic indexers dealt with distinct disciplines and, as long as they did not stray beyond disciplinary boundaries, a quick and dirty keyword approach was satisfactory. The trouble came when attempts were made to make a combined index for two or more distinct disciplines. A number of processes have since been developed, mostly involving tagging of some kind or use of strings. Mooers' solution has rarely been considered seriously and probably would be extremely difficult to apply now because of so much interdisciplinarity. But for a specific, weIl defined field, it is still weIl worth considering. Mooers received training in mathematics and physics from the University of Minnesota and the Massachusetts Institute of Technology. He was the founder of Zator Company, which developed and marketed a coded card information retrieval system, and of Rockford Research, Inc., which engages in research in information science. He is the inventor of the TRAC computer language.

Footnote

Original in: Information retrieval today: papers presented at an Institute conducted by the Library School and the Center for Continuation Study, University of Minnesota, Sept. 19-22, 1962. Ed. by Wesley Simonton. Minneapolis, Minn.: The Center, 1963. S.21-36.

0.013245909 = product of:
  0.026491817 = sum of:
    0.026491817 = product of:
      0.07947545 = sum of:
        0.07947545 = weight(_text_:themes in 3642) [ClassicSimilarity], result of:
          0.07947545 = score(doc=3642,freq=2.0), product of:
            0.3729592 = queryWeight, product of:
              6.429029 = idf(docFreq=193, maxDocs=44218)
              0.058011748 = queryNorm
            0.21309422 = fieldWeight in 3642, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              6.429029 = idf(docFreq=193, maxDocs=44218)
              0.0234375 = fieldNorm(doc=3642)
      0.33333334 = coord(1/3)
  0.5 = coord(1/2)

Abstract

This chapter from The Sayers Memorial Volume summarizes Farradane's earlier work in which he developed his major themes by drawing in part upon research in psychology, and particularly those discoveries called "cognitive" which now form part of cognitive science. Farradane, a chemist by training who later became an information scientist and Director of the Center for Information Science, City University, London, from 1958 to 1973, defines the various types of methods used to achieve classification systems-philosophic, scientific, and synthetic. Early an he distinguishes the view that classification is "some part of external 'reality' waiting to be discovered" from that view which considers it "an intellectual operation upon mental entities and concepts." Classification, therefore, is to be treated as a mental construct and not as something "out there" to be discovered as, say, in astronomy or botany. His approach could be termed, somewhat facetiously, as an "in there" one, meaning found by utilizing the human brain as the key tool. This is not to say that discoveries in astronomy or botany do not require the use of the brain as a key tool. It is merely that the "material" worked upon by this tool is presented to it for observation by "that inward eye," by memory and by inference rather than by planned physical observation, memory, and inference. This distinction could be refined or clarified by considering the initial "observation" as a specific kind of mental set required in each case. Farradane then proceeds to demolish the notion of main classes as "fictitious," partly because the various category-defining methodologies used in library classification are "randomly mixed." The implication, probably correct, is that this results in mixed metaphorical concepts. It is an interesting contrast to the approach of Julia Pettee (q.v.), who began with indexing terms and, in studying relationships between terms, discovered hidden hierarchies both between the terms themselves and between the cross-references leading from one term or set of terms to another. One is tempted to ask two questions: "Is hierarchy innate but misinterpreted?" and "ls it possible to have meaningful terms which have only categorical relationships (that have no see also or equivalent relationships to other, out-of-category terms)?" Partly as a result of the rejection of existing general library classification systems, the Classification Research Group-of which Farradane was a charter member decided to adopt the principles of Ranganathan's faceted classification system, while rejecting his limit an the number of fundamental categories. The advantage of the faceted method is that it is created by inductive, rather than deductive, methods. It can be altered more readily to keep up with changes in and additions to the knowledge base in a subject without having to re-do the major schedules. In 1961, when Farradane's paper appeared, the computer was beginning to be viewed as a tool for solving all information retrieval problems. He tartly remarks: