Search (2 results, page 1 of 1)

0.0074047255 = product of:
  0.059237804 = sum of:
    0.059237804 = weight(_text_:case in 4383) [ClassicSimilarity], result of:
      0.059237804 = score(doc=4383,freq=2.0), product of:
        0.1742197 = queryWeight, product of:
          4.3964143 = idf(docFreq=1480, maxDocs=44218)
          0.03962768 = queryNorm
        0.34001783 = fieldWeight in 4383, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          4.3964143 = idf(docFreq=1480, maxDocs=44218)
          0.0546875 = fieldNorm(doc=4383)
  0.125 = coord(1/8)

Abstract

To investigate the extent to which online catalogs arrange together, or collocate, records representing particular authors and works, a survey compared the displays resulting from 5 author and 5 work queries in 18 online catalogs. Dependent variables to measure collocation included the number of items irrelevant records were interfiled among relevant records. Searches for worst-case authors and works associated with large retrieval sets, including 'Homer' and 'Paradise lost', revealed the effects of Boolean versus string matching, query type, and catalog size on the collocation of relevant records. Results of the survey showed that string matching collocated relevant records more successfully than Boolean matching, that author records were collocated more successfully than work records, and, surprisingly, that catalog size had only a small effect on collocation

0.00522842 = product of:
  0.04182736 = sum of:
    0.04182736 = weight(_text_:studies in 449) [ClassicSimilarity], result of:
      0.04182736 = score(doc=449,freq=2.0), product of:
        0.15812531 = queryWeight, product of:
          3.9902744 = idf(docFreq=2222, maxDocs=44218)
          0.03962768 = queryNorm
        0.26452032 = fieldWeight in 449, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.9902744 = idf(docFreq=2222, maxDocs=44218)
          0.046875 = fieldNorm(doc=449)
  0.125 = coord(1/8)

Abstract

Central to subject searching is the match between user vocabulary and the headings from Library of Congress Subject-Headings (LCSH) used in a library catalog. This paper evaluates previous matching studies, proposes a detailed list of matching categories, and tests LCSH in a study using these categories. Exact and partial match categories are defined for single LCSH and multiple LCSH matches to user expressions. One no-match category is included. Transaction logs from ORION, UCLA's online Information system, were used to collect user expressions for a comparison of LCSH and user language. Results show that single LCSH headings match user expressions exactly about 47% of the time; that single subject heading matches, including exact matches, comprise 74% of the total; that partial matches, to both single and multiple headings, comprise about 21% of the total; and that no match occurs 5% of the time.