Hyperlink Analysis

Abstract

Hyperlink analysis is a collection of techniques that researchers can use to identify patterns in the hyperlinks between collections of web sites. The purpose can either be to investigate linking behaviour within a given community or to use links as a convenient data source with which to investigate an aspect of or a reflection of a wider behaviour type. An example of a direct application is investigating the types of links used in political web sites (Foot, Schneider, Dougherty, Xenos, & Larsen, 2003; Park, 2003). Examples of indirect applications include using links between university web sites to identify highly connected pairs of sites (Thelwall, 2002b), and using links to access whether the web publishing of highly-rated scholars is better than that of others (Thelwall & Harries, 2004). There are two general approaches used in hyperlink analysis. The first is a content analysis of the hyperlinks themselves, categorising them by their context in their source web page and using the categorisation results for a descriptive analysis of why hyperlinks are used in the chosen set of web sites. The second general approach for hyperlink analysis is to choose a set of web sites relevant to a research question and then count links between all pairs of sites within the set, then applying statistical techniques to identify or verify a pattern in the link counts.