Search (69 results, page 4 of 4)

0.004806533 = product of:
  0.03364573 = sum of:
    0.016822865 = weight(_text_:classification in 2804) [ClassicSimilarity], result of:
      0.016822865 = score(doc=2804,freq=2.0), product of:
        0.09562149 = queryWeight, product of:
          3.1847067 = idf(docFreq=4974, maxDocs=44218)
          0.03002521 = queryNorm
        0.17593184 = fieldWeight in 2804, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.1847067 = idf(docFreq=4974, maxDocs=44218)
          0.0390625 = fieldNorm(doc=2804)
    0.016822865 = weight(_text_:classification in 2804) [ClassicSimilarity], result of:
      0.016822865 = score(doc=2804,freq=2.0), product of:
        0.09562149 = queryWeight, product of:
          3.1847067 = idf(docFreq=4974, maxDocs=44218)
          0.03002521 = queryNorm
        0.17593184 = fieldWeight in 2804, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.1847067 = idf(docFreq=4974, maxDocs=44218)
          0.0390625 = fieldNorm(doc=2804)
  0.14285715 = coord(2/14)

Abstract

With the increasing number of digital documents, the ability to automatically classify those documents both efficiently and accurately is becoming more critical and difficult. One of the major problems in text classification is the high dimensionality of feature space. We present the ambiguity measure (AM) feature-selection algorithm, which selects the most unambiguous features from the feature set. Unambiguous features are those features whose presence in a document indicate a strong degree of confidence that a document belongs to only one specific category. We apply AM feature selection on a naïve Bayes text classifier. We favorably show the effectiveness of our approach in outperforming eight existing feature-selection methods, using five benchmark datasets with a statistical significance of at least 95% confidence. The support vector machine (SVM) text classifier is shown to perform consistently better than the naïve Bayes text classifier. The drawback, however, is the time complexity in training a model. We further explore the effect of using the AM feature-selection method on an SVM text classifier. Our results indicate that the training time for the SVM algorithm can be reduced by more than 50%, while still improving the accuracy of the text classifier. We favorably show the effectiveness of our approach by demonstrating that it statistically significantly (99% confidence) outperforms eight existing feature-selection methods using four standard benchmark datasets.

0.0045916047 = product of:
  0.03214123 = sum of:
    0.024005229 = weight(_text_:subject in 2741) [ClassicSimilarity], result of:
      0.024005229 = score(doc=2741,freq=4.0), product of:
        0.10738805 = queryWeight, product of:
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03002521 = queryNorm
        0.22353725 = fieldWeight in 2741, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03125 = fieldNorm(doc=2741)
    0.008136002 = product of:
      0.016272005 = sum of:
        0.016272005 = weight(_text_:22 in 2741) [ClassicSimilarity], result of:
          0.016272005 = score(doc=2741,freq=2.0), product of:
            0.10514317 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03002521 = queryNorm
            0.15476047 = fieldWeight in 2741, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03125 = fieldNorm(doc=2741)
      0.5 = coord(1/2)
  0.14285715 = coord(2/14)

Abstract

This study seeks to find out how human beings cluster Web pages naturally. Twenty Web pages retrieved by the Northem Light search engine for each of 10 queries were sorted by 3 subjects into categories that were natural or meaningful to them. lt was found that different subjects clustered the same set of Web pages quite differently and created different categories. The average inter-subject similarity of the clusters created was a low 0.27. Subjects created an average of 5.4 clusters for each sorting. The categories constructed can be divided into 10 types. About 1/3 of the categories created were topical. Another 20% of the categories relate to the degree of relevance or usefulness. The rest of the categories were subject-independent categories such as format, purpose, authoritativeness and direction to other sources. The authors plan to develop automatic methods for categorizing Web pages using the common categories created by the subjects. lt is hoped that the techniques developed can be used by Web search engines to automatically organize Web pages retrieved into categories that are natural to users. 1. Introduction The World Wide Web is an increasingly important source of information for people globally because of its ease of access, the ease of publishing, its ability to transcend geographic and national boundaries, its flexibility and heterogeneity and its dynamic nature. However, Web users also find it increasingly difficult to locate relevant and useful information in this vast information storehouse. Web search engines, despite their scope and power, appear to be quite ineffective. They retrieve too many pages, and though they attempt to rank retrieved pages in order of probable relevance, often the relevant documents do not appear in the top-ranked 10 or 20 documents displayed. Several studies have found that users do not know how to use the advanced features of Web search engines, and do not know how to formulate and re-formulate queries. Users also typically exert minimal effort in performing, evaluating and refining their searches, and are unwilling to scan more than 10 or 20 items retrieved (Jansen, Spink, Bateman & Saracevic, 1998). This suggests that the conventional ranked-list display of search results does not satisfy user requirements, and that better ways of presenting and summarizing search results have to be developed. One promising approach is to group retrieved pages into clusters or categories to allow users to navigate immediately to the "promising" clusters where the most useful Web pages are likely to be located. This approach has been adopted by a number of search engines (notably Northem Light) and search agents.

Date

12. 9.2004 9:56:22

0.0036373418 = product of:
  0.05092278 = sum of:
    0.05092278 = weight(_text_:subject in 507) [ClassicSimilarity], result of:
      0.05092278 = score(doc=507,freq=2.0), product of:
        0.10738805 = queryWeight, product of:
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03002521 = queryNorm
        0.4741941 = fieldWeight in 507, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.09375 = fieldNorm(doc=507)
  0.071428575 = coord(1/14)

0.0036373418 = product of:
  0.05092278 = sum of:
    0.05092278 = weight(_text_:subject in 1043) [ClassicSimilarity], result of:
      0.05092278 = score(doc=1043,freq=2.0), product of:
        0.10738805 = queryWeight, product of:
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03002521 = queryNorm
        0.4741941 = fieldWeight in 1043, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.09375 = fieldNorm(doc=1043)
  0.071428575 = coord(1/14)

0.0017434291 = product of:
  0.024408007 = sum of:
    0.024408007 = product of:
      0.048816014 = sum of:
        0.048816014 = weight(_text_:22 in 1046) [ClassicSimilarity], result of:
          0.048816014 = score(doc=1046,freq=2.0), product of:
            0.10514317 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03002521 = queryNorm
            0.46428138 = fieldWeight in 1046, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.09375 = fieldNorm(doc=1046)
      0.5 = coord(1/2)
  0.071428575 = coord(1/14)

Date

5. 5.2003 14:17:22

0.0015155592 = product of:
  0.021217827 = sum of:
    0.021217827 = weight(_text_:subject in 5702) [ClassicSimilarity], result of:
      0.021217827 = score(doc=5702,freq=2.0), product of:
        0.10738805 = queryWeight, product of:
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03002521 = queryNorm
        0.19758089 = fieldWeight in 5702, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.0390625 = fieldNorm(doc=5702)
  0.071428575 = coord(1/14)

Abstract

In this article, we evaluate the retrieval performance of an algorithm that automatically categorizes medical documents. The categorization, which consists in assigning an International Code of Disease (ICD) to the medical document under examination, is based on wellknown information retrieval techniques. The algorithm, which we proposed, operates in a fully automatic mode and requires no supervision or training data. Using a database of 20,569 documents, we verify that the algorithm attains levels of average precision in the 70-80% range for category coding and in the 60-70% range for subcategory coding. We also carefully analyze the case of those documents whose categorization is not in accordance with the one provided by the human specialists. The vast majority of them represent cases that can only be fully categorized with the assistance of a human subject (because, for instance, they require specific knowledge of a given pathology). For a slim fraction of all documents (0.77% for category coding and 1.4% for subcategory coding), the algorithm makes assignments that are clearly incorrect. However, this fraction corresponds to only one-fourth of the mistakes made by the human specialists

0.0015155592 = product of:
  0.021217827 = sum of:
    0.021217827 = weight(_text_:subject in 3300) [ClassicSimilarity], result of:
      0.021217827 = score(doc=3300,freq=2.0), product of:
        0.10738805 = queryWeight, product of:
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.03002521 = queryNorm
        0.19758089 = fieldWeight in 3300, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          3.576596 = idf(docFreq=3361, maxDocs=44218)
          0.0390625 = fieldNorm(doc=3300)
  0.071428575 = coord(1/14)

Abstract

Automatic document categorization is an important research problem in Information Science and Natural Language Processing. Many applications, including, Word Sense Disambiguation and Information Retrieval in large collections, can benefit from such categorization. This paper focuses on automatic categorization of documents from the biomedical literature into broad discipline-based categories. Two different systems are described and contrasted: CISMeF, which uses rules based on human indexing of the documents by the Medical Subject Headings (MeSH) controlled vocabulary in order to assign metaterms (MTs), and Journal Descriptor Indexing (JDI), based on human categorization of about 4,000 journals and statistical associations between journal descriptors (JDs) and textwords in the documents. We evaluate and compare the performance of these systems against a gold standard of humanly assigned categories for 100 MEDLINE documents, using six measures selected from trec_eval. The results show that for five of the measures performance is comparable, and for one measure JDI is superior. We conclude that these results favor JDI, given the significantly greater intellectual overhead involved in human indexing and maintaining a rule base for mapping MeSH terms to MTs. We also note a JDI method that associates JDs with MeSH indexing rather than textwords, and it may be worthwhile to investigate whether this JDI method (statistical) and CISMeF (rule-based) might be combined and then evaluated showing they are complementary to one another.

0.0014528577 = product of:
  0.020340007 = sum of:
    0.020340007 = product of:
      0.040680014 = sum of:
        0.040680014 = weight(_text_:22 in 611) [ClassicSimilarity], result of:
          0.040680014 = score(doc=611,freq=2.0), product of:
            0.10514317 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03002521 = queryNorm
            0.38690117 = fieldWeight in 611, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.078125 = fieldNorm(doc=611)
      0.5 = coord(1/2)
  0.071428575 = coord(1/14)

Date

22. 8.2009 12:54:24

5.811431E-4 = product of:
  0.008136002 = sum of:
    0.008136002 = product of:
      0.016272005 = sum of:
        0.016272005 = weight(_text_:22 in 3284) [ClassicSimilarity], result of:
          0.016272005 = score(doc=3284,freq=2.0), product of:
            0.10514317 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03002521 = queryNorm
            0.15476047 = fieldWeight in 3284, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.03125 = fieldNorm(doc=3284)
      0.5 = coord(1/2)
  0.071428575 = coord(1/14)

Date

22. 1.2010 14:41:24