MEASURING THE WWW - OCLC�S IMPORTANT RESEARCH PROJECT

Patricia Layzell Ward, Editor, Library Link

One of the questions that could be posed at quiz nights is "How many sites are there on the World Wide Web?" Visiting OCLC�s website and looking at the results obtained from its Web Characterization Project can provide the answer. The latest statistics are dated June 1999. The web has about 3.6 million sites, of which 2.2 million are publicly accessible. Researchers found that the largest 25,000 sites represent about 50% of the Web�s content, and that the number of sites and their size are climbing. Another question to pose might be "How many pages are there on the 2.2 publicly accessible sites?" The answer is nearly 300 million Web pages.

The project has also revealed that significant portions of the Web are not publicly accessible or do not offer meaningful content. About 2 % of the public sites - 42,000 of the 2.2 million - contain sexually explicit material.

The mean size of a public Web site is about 129 pages, a 13% increase over last year�s estimate of 114 pages. The number of public Web sites has nearly tripled in the two-year period from June 1997 to June 1999, increasing from 800,000 to 2.2 million. An indication of Web volatility is that of the IP addresses identifying a Web site in 1998 that no longer identify a Web site in 1999 is 44%.

The project is based in the OCLC Office of Research, and the Web Characterization Project is an ongoing initiative. It focuses on the structure, size, usage, and content of the Web. It is also concerned with developing data collection methodologies and metrics for Web description. The project is managed by Dr. Ed O�Neill.

The OCLC Web site carries fuller statistics that are regularly updated together with an explanation of the statistics, and more details about the project. Papers relating to the Web Characterization Repository and W3C Characterizatrion Activity are available through linkages on the OCLC site at http://www.oclc.org/oclc/research/projects/webstats

Students looking for reliable statistics about the Web or topics for assignments will find the site repays a visit. Likewise librarians and information managers needing authoritative data for reports can find them here. OCLC is to be congratulated, not only for sponsoring the research, but also for effectively disseminating the results.

Back to Electronic Resources Viewpoints