Hilltop algorithm

from Wikipedia, the free encyclopedia

The Hilltop algorithm ( Hilltop dt. "Bergspitze") is a method that sorts a large number of linked documents (for example the World Wide Web ) according to relevance with regard to search terms. In contrast to PageRank - algorithm so not the value is determined of a document, but its connection to a particular search term .

history

The algorithm was developed by Krishna Bharat and George A. Mihaila at the University of Toronto . In 2003, Google acquired the patent.

procedure

The Hilltop algorithm consists of two successive phases:

  1. Determination of the so-called expert pages for a search term. In the Hilltop algorithm, documents that refer to a large number of independent documents on this topic are defined as expert pages. These expert pages can be determined automatically.
  2. Sorting of the authority pages based on the frequency and relevance of the referring expert pages. By definition, authority pages are documents to which at least two independent expert pages refer.

Web links