Determining Authoritiveness on the Web

Abstract Assessments of the accuracy or trustworthiness of information in a computer system fall within the scope of data and communications security. For the purposes of this article, we define “security” very broadly, to include the provision of promised information, of ascertainable quality, intact and unaltered, to a desired audience, across appropriate ranges of time and space. In traditional media-radio, television, and print-editors and publishers pass judgement upon authors' works before they reach the public. With the advent of the World Wide Web, authors can present their work without editorial intervention. In such cases, readers must judge works based on internal consistencies, citations, methods, and upon the credentials provided by the authors themselves. In this article, a selection of academics' personal Web pages is discussed from a Soft Systems perspective, with the aid of content analysis and iconographic examinations. Materials studied were freely available to the public. The contents of academics' personal Web pages were examined with regard to the traditional assessment factors: teaching, research, and service, and compared to the stated or inferred career stages or academic ranks of the authors. The types of materials presented-curricular, personal information, and research works-did not appear to be directly linked to academic rank. This suggests that users are apt to receive information both from established scholars and from academics whose careers are relatively new. While this in itself is arguably beneficial, it places an extra onus upon intermediaries and end users, who must judge for themselves which points of view dominate in a given field. This investigation suggests that the Web may present a problem situation with which the dominant librarian's model for assessing scholarly literature may be insufficient. We posit the need for a far looser model, employing techniques from other social sciences and humanities.