Hyperlink Structure of Electronic Commerce Websites

Web Structure Mining (WSM) Research

WSM represents mining the hyperlinks between inter and intra Websites to understand the overall structure to extract patterns in the information architecture. Exploitation of hyperlink structure of the Web allows the identification of interesting information, such as the patterns linking the Web pages, and Websites, ranking and classifying of Web pages based on hyperlinks (Lappas, 2007). Hyperlink structure analysis of the Web is used as a basis to obtain search results (Espadas, Calero, & Piattini, 2008). Search engines use the hyperlinks to identify more relevant Web pages by ranking based on the number of links pointed to a Web page (Kosala & Blockeel, 2000). One of the well-known examples is the commercial search engine Google’s, page rank algorithm which uses hyperlinks to rank the Web pages for a specific keyword search (Brin & Page, 1998). The Web page that receives more incoming links relatively is ranked higher than other similar Web pages.

Hyperlink structure has been studied and used in information retrieval algorithms to rank search results on the Web (Vaughan & You, 2005; L. Yan, Wei, Gui, & Chen, 2011). Statistical analysis of Web page links has been conducted by (Becchetti, Castillo, Donato, Baeza-Yates, & Leonardi, 2008) and they proposed spam detection techniques that are only based on the link structure of the Web page irrespective of the contents. Some studies have exploited commerce domain Websites’ hyperlinks and found that the number of links pointing to a commerce company’s Website is correlated with the company’s business information such as revenue, profit and research expenses (Romero-Frias & Vaughan, 2010; Vaughan & Wu, 2004; Vaughan, 2004a). The findings of these research works suggest that hyperlinks to commercial Websites could be used as a business performance indicator. This knowledge is useful for competitive business intelligence and Web data mining.