Article Preview
TopIntroduction
The Semantic Web can be regarded as a smart web designed to understand the requests of people and machines to use Web contents (Berners-Lee, 2001), and one of the most common requests may be the ranking of Semantic Web resources. Therefore the Web had better have some understanding of the concept of ranking, and the ranking mechanism adequate for a specific domain could be defined in the domain ontology. However, there have not been many studies on ranking in the Semantic Web, while there have been extensive studies on the evaluation of the World Wide Web.
Traditionally, the importance of a particular Web page is estimated based on the number of keywords found on the page, which is subject to manipulation (Marchiori, 1997). In contrast, link analysis methods such as Google’s PageRank (Brin et al., 1998; Haveliwala, 1999; Page et al., 1998) capitalize on the information that is inherent in the link structure of a Web graph. PageRank considers a page important if it is referred to by many other pages. The degree of importance also increases if the importance of the referring pages is high. Kleinberg’s Hypertext-Induced Topic Selection (HITS) algorithm (Kleinberg, 1998) is another link-structure-based ranking algorithm for Web pages. The HITS algorithm differs from PageRank in that it utilizes two kinds of scores: an authority score and a hub score. If a page has a high authority score, it is an authority on a given topic and many pages refer to it. A page with a high hub score links to many authoritative pages. The link-structure-based ranking method has become an essential tool for using WWW, and its effectiveness and efficiency have been widely recognized.
On the other hand, information from the Semantic Web can be expressed using a Resource Description Framework (RDF) graph (Klyne et al., 2004; Manola et al., 2004). An RDF graph, in which a resource and a property are expressed as a node and a link, respectively, is similar to a WWW graph in which a Web page and a hyperlink are expressed as a node and a link, respectively. Consequently, research on methods for applying the link-structure-based ranking technique of WWW to an RDF graph of the Semantic Web has great significance. The WWW graph can be thought of as an enormous class of the Web pages with only one recursive property called a ‘refer to’ property, so to speak. An RDF schema, in contrast, can have various classes and properties, and each link corresponding to a property can have an opposite direction depending on whether it is an active or a passive voice. As a result, RDF schemas can have many different forms because the direction of each link is changeable, even if they describe the same thing, and the direction of an RDF link does not have the same meaning as that of a WWW hyperlink. In WWW, if a page is pointed to by a directing link, we can tell that the pointed page must have some useful information. PageRank and HITS are based on this basic assumption.