C. Castillo. School of Engineering, Santiago, Chile, (November 2004)
The key factors for the success of the World Wide Web are its large size and the lack of a centralized control over its contents. Both issues are also the most important source of problems for locating information. The Web is a context in which traditional Information Retrieval methods are challenged, and given the volume of the Web and its speed of change, the coverage of modern search engines is relatively small. Moreover, the distribution of quality is very skewed, and interesting pages are scarce in comparison with the rest of the content. Web crawling is the process used by search engines to collect pages from the Web. This thesis studies Web crawling at several different levels, ranging from the long-term goal of crawling important pages first, to the short-term goal of using the network connectivity efficiently, including implementation issues that are essential for crawling in practice. We start by designing a new model and architecture for aWeb crawler that tightly integrates the crawler with the rest of the search engine, providing access to the metadata and links of the documents that can be used to guide the crawling process effectively. We implement this design in the WIRE project as an efficient Web crawler that provides an experimental framework for this research. In fact, we have used our crawler to characterize the Chilean Web, using the results as feedback to improve the crawler design. We argue that the number of pages on the Web can be considered infinite, and given that a Web crawler cannot download all the pages, it is important to capture the most important ones as early as possible during the crawling process. We propose, study, and implement algorithms for achieving this goal, showing that we can crawl 50% of a large Web collection and capture 80% of its total Pagerank value in both simulated and real Web environments. We also model and study user browsing behavior in Web sites, concluding that it is not necessary to go deeper than five levels from the home page to capture most of the pages actually visited by people, and support this conclusion with log analysis of several Web sites. We also propose several mechanisms for server cooperation to reduce network traffic and improve the representation of aWeb page in a search engine with the help of Web site managers.
