Augmented Conceptual Analysis of the Web

ABSTRACT

Keywords

© 1997 Copyright on this material is held by the authors.

INTRODUCTION

The analyses we seek to develop in the workshop must be grounded in observable phenomena. For example, there has recently been an observable antagonism between content providers and indexing services on the Web. Search engines, fed by crawler technology, have become a dominant means for users to discover new information of interest on the Web. The advantages to content providers of being listed at or near to the top of a user's search results are enormous: good positioning in the list confers the power to attract users' attention and time. As a result, creators of Web pages can be observed to deliberately strategize to maximize their "score" with the crawlers and search engines. Their strategies have ranged from the subtle, for example, naming pages to occur first in alphabetized lists, to blatant pandering, for example, repeating likely search keywords in a paragraph at the bottom of the page. In response, crawlers and search engines have begun to be more sophisticated in how they assess a page's content -- for example, ignoring or discounting repeated (in a row) instances of a keyword.

The example above might be understood within an evolutionary perspective on analogy to a scenario of co-evolution (e.g., of predator and prey), or by an info rmation-theoretic analysis of the cost structure of discovering information, such as Pirolli and Card's (1995) analysis of information foraging. Applying a perspective or an analogy to an observed phenomenon can increase the level of abstraction at which it is characterized. The workshop will seek to exploit this by developing an inventory of significant Web phenomena, and investigating a variety of perspectives, such as evolutionary analysis, to abstract away from the details of particular examples. Useful perspectives may be historical, empirical, social, political, biological, information-theoretic, technological, human-computer interaction, or evolutionary. Some of these will no doubt prove more useful than others in provoking insight. One outcome of the workshop will be an understanding of which abstractions are productive.

We note that there are several unique features that make the Web an interesting domain for analysis:

• Unprecedented rapid growth and change.
• The Web's reach to more people than ever (by the time of the workshop there may be 2 million web sites and thus on the order of 2 million user interface designers!).
• Macro-CSCW. Millions of users cooperating via the web suggests the need for an analysis that contrasts ad hoc collaboration from the intentional collaboration that is the signature of traditional CSCW.
• Cross-platform. What are the tradeoffs for developers and users?
• Slow-link client-servers. People's patience, rather than technical feasibility, has become the upper limit on design.
• Authoring is getting harder. HTML is no longer enough for leading-edge Web design. The schism between professional and amateur content providers, and those with and without technical know-how is widening.
• Proxies. Proxies offer the possibility of augmenting, or radically altering information passed between clients and servers. This possibility touches on a number of potent issues, e.g., intellectual property, privacy, derivative works, etc.
• Evolution of protocols. New protocols are emerging, creating an ever-changing landscape for the Web.
• E-commerce. As commerce grows in importance, many new concepts for buying and selling goods, driven by technological innovation, will emerge.

REFERENCE

Workshop Report

Workshop Organizers

Jakob Nielsen is a Sun Microsystems Distinguished Engineer working on advanced Web technology at SunSoft, Sun's software planet. He was Papers co-chair for INTERCHI'93 and has served in a variety of other CHI and SIGCHI positions. Dr. Nielsen has chaired or co-chaired workshops at CHI+GI'87, CHI'88, CHI'92, and CHI'94 and participated in several additional CHI workshops. He writes the monthly Alertbox column on Web usability.