Formal methods for verification of websites macrostructure integrity

The problem of verifying macrostructure of websites where information resources are represented by means of HTML is studied. The macrostructure of a website is understood as the structure of interrelations between its resources based on hyperlinks. The approach presented involves inventorying information resources of the site, partial syntactic analysis of its HTML-files, locating their hyperlinks, and, in so doing, reconstructing its macrostructure (the hyperlink graph of resources). The description of the site's macrostructure is placed into a relational database and is analyzed by formal methods based on relational algebra and graph theory in order to detect irrelevant (dangling) hyperlinks, resources to which no inner hyperlink points, as well as hypertext pages unattainable from the homepage of the site. In the paper, the prospects of this approach are discussed in relation to websites created with the help of XML technologies. A brief description of an implemented prototype of a verification system for HTML sites is given.