Similarity and Ranking Operations
Association; Correlation; Matching; Ordering; Proximity; Relevance
Similarity and ranking operations are fundamental to information searching, in which a user generates a query phrase of one or more words that reflects an information need. The query is used to find related items that satisfy that need.
Similarity operations quantify the resemblance or alikeness between two information objects. An information object is a conceptual unit, most typically described as a document, but also taking the form of a term, phrase, paragraph, page, section, chapter, article, book, or script, etc. Information objects may be printed or digital.
Similarity judgments may be subjective (performed by a user) or algorithmic (performed by a computer). Depending on the method of evaluation, alikeness judgments may be semantic (i.e., within the meaning of a document), structural (i.e., within parts of speech, position in a document, or pattern of links between documents), or statistical...
- 1.Harman D. Ranking algorithms. In: Frakes WB, Baeza-Yates R, editors. Information retrieval: data structures & algorithms, chap. 14. Upper Saddle River: Prentice-Hall; 1992. p. 363–92.Google Scholar
- 3.Rasmussen EM. Clustering algorithms. In: Frakes WB, Baeza-Yates R, editors. Information retrieval: data structures & algorithms, chap. 16. Upper Saddle River: Prentice-Hall; 1992. p. 419–42.Google Scholar