Informetric theories and methods for exploring the Internet: an analytical survey of recent research literature

Library Trends, Wntr, 2002 by Judit Bar-Ilan, Bluma C. Peritz

Not only discussion lists were analyzed, but also Web pages and Web sites. Cronin et al. (1998) searched the Web using five search tools for pages mentioning five prominent professors in library and information science. The retrieved Web pages were characterized according to the "forms of mention." Eleven categories of invocation were defined: Abstract, article, conference proceedings, current awareness, external home page, listserv, personal/parent organization home page, resource guide, book review, syllabus, and table of contents. The data were collected over a period of two months, though the dates are not given. The authors concluded: The Web "engenders new modes of scholarly interaction and signaling. Scholars do not merely post, or publish, their works on the Web: They seed ideas, discuss issues, and debate positions, in ways which, occasionally, deviate from, and challenge, established norms" (p. 1326).

A different kind of content analysis, examining not the form of invocation, but the different contexts in which the mathematician Paul Erdos was mentioned, appears in Bar-Ilan (1998). The paper analyzes the content of 2,685 Web documents collected between the end of 1996 and the beginning of 1997 (Paul Erdos passed away in September, 1996). Six main content categories were defined: Mathematical work, Erdos number, in honor/memory of Erdos,jokes/quotations, math education, and other. Almost 40 percent of the pages were classified as "mathematical work," but a rather surprising 13 percent of the pages belonged to the jokes/quotations category. (The most popular quotations/jokes were: "A mathematician is a machine for turning coffee into theorems" and "Why did the chicken cross the road? It was forced to do so by the chicken-hole principle"). The concept of Erdos number intrigues the authors of the Web pages; the concept was explained on ninety-one (3 percent) different pages (almost always exactly the same explanation), and 9 percent of the collected pages point to the home page of the "Erdos Number Project" (http://www.oakland.edu/~grossman/erdoshp.html). In 40 percent of the pages belonging to "mathematical work," Erdos's name was mentioned in bibliographical references.

Formal bibliographical references also appeared in Bar-Ilan (2000c), in a large portion of the pages (in 40.3 percent out of the 807 pages) containing the search terms "informetrics OR informetric." The searches were carried out in June 1998 using the six largest search engines at that time (Alta Vista, Excite, Hotbot, Infoseek, Lycos, and Northern Light). The references extracted from these pages (called the "Web database") were compared with comparable data retrieved from commercial bibliographical databases. In all except one comparison, the Web database did at least as well as the commercial database, indicating that valuable, freely available data exist in the Web, but cannot be located easily.

Lawrence, Bollacker, & Giles (1999) were able to find large quantities of full-text papers in the area of computer science. They were looking for different formats, including PDF and PostScript. Bar-Ilan's findings were rather different; she located only a negligible number (4) of full text publications. This may be due to the fact that she collected information about a different subject area or to the fact that general search engines ignore formats like PDF and PostScript. The most productive and the most cited authors and sources, and the most cited papers (papers which are referred to in the largest number of collected Web pages) were also calculated.