Beyond relevance ranking: hyperlink vector voting
Yanhong Li, Larry Rafsky · 1997
A new method for hypertext indexing and retrieval called Hyperlink Vector Voting (HVV) is proposed. It combines relevance ranking and quality ranking for hypertext retrieval systems. Ranking of search results no longer depends on the content of the document being ranked but rather on the number of hyperlinks pointing to the document and the descriptions of these links. An experimental World Wide Web search engine is also described, and the results it produces appear quite satisfactory.