Open AccessRanking the Online Documents Based on Relative Credibility Measures
Author(s) -
Ahmad Dahlan,
Dwi H. Widyantoro,
Suhono Harso Supangkat,
Benhard Sitohang
Publication year - 2009
Publication title -
itb journal of information and communication technology
Language(s) - English
Resource type - Journals
ISSN - 1978-3086
DOI - 10.5614/itbj.ict.2009.3.1.2
Subject(s) - ranking (information retrieval) , credibility , information retrieval , pagerank , relevance (law) , rank (graph theory) , computer science , rank correlation , spearman's rank correlation coefficient , search engine , metasearch engine , learning to rank , citation , mean reciprocal rank , the internet , data mining , mathematics , world wide web , machine learning , web search query , political science , law , combinatorics
Information searching is the most popular activity in Internet. Usually the search engine provides the search results ranked by the relevance. However, for a certain purpose that concerns with information credibility, particularly citing information for scientific works, another approach of ranking the search engine results is required. This paper presents a study on developing a new ranking method based on the credibility of information. The method is built up upon two well-known algorithms, PageRank and Citation Analysis. The result of the experiment that used Spearman Rank Correlation Coefficient to compare the proposed rank (generated by the method) with the standard rank (generated manually by a group of experts) showed that the average Spearman 0 < rS < critical value. It means that the correlation was proven but it was not significant. Hence the proposed rank does not satisfy the standard but the performance could be improved
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom