Ph.D. Thesis - Forward, Back and Home Again
Analyzing User Behavior on the Web

Propositions supplemental to the Ph.D. thesis

1

There is no such thing as a `typical Web user' or a `typical Web navigation session'. Hence, rather than to strive toward one integrated theory of Web navigation, it makes more sense to strive toward a comprehensive body of empirical data and theoretical insights that each provide different views on Web navigation.

2

The development and evaluation of relatively straightforward, context-aware support concepts is an issue that deserves more attention from the adaptive hypermedia community.

3

The concept of aggregated navigation styles reflects more detailed aspects than a battery of first-order complexity navigation measures. The subjectivity introduced by a careful interpretation step is preferable to the various possible interpretations of each single first-order measure.

4

From a computer science point of view, discovering user needs and representing them in a suitable format is the most challenging part of the adaptation process; the actual adaptation decision may be regarded as a mere user interface design issue.

5

Inference of long-term user interests based on the users' general Web activities will result in overly general profiles that only cover a limited part of the users' interests. However, as most Web activities take place in longer navigation sessions, in most situations the current user interests can be estimated from the activities earlier in the session.

6

The evolution of the Web from a static hypermedia system to a hybrid between hypermedia and interactive tools and applications requires a major rethinking of the metaphor of Web navigation into an interleaving process of navigation and interaction activities.

7

In particular for pages that are visited on an infrequent or irregular basis, it is less likely that a user will remember the address. It is ironic yet unavoidable that browsers cannot provide a list of possible items that is helpful in these situations.

8

Vanevar Bush' original concept of trails should be reintroduced in the Web interface, to provide users with orientation support for backtracking activities, and to help users in relocating infrequently or irregularly visited pages.

9

Orienteering behavior, the combined process of search and exploration of information patches, allows users to find what they need rather than what they asked for. For this reason, hyperlink navigation will remain one of the Web's main interaction paradigms.

10

Usability research is the art of explaining why users should be annoyed by certain aspects of the user interface.


Hannover, Eelco Herder
March 20, 2006