Usability and the Semantic Web
By: Anthony Jameson (2006)
In Y. Sure & J. B. Domingue (Eds.), The Semantic Web: Research and Applications: Proceedings of the Third European Semantic Web Conference, ESWC 2006 (p. 3). Berlin: Springer. Abstract of an invited talk.
Note: The presentation slides (see the link below) are much
more detailed than the brief abstract in the proceedings
volume. The slides contain a good deal of supplementary
text that make the slides reasonably self-contained,
including the following summary:
Many innovative technologies aim to enhance a user’s interaction with a system
in some respects; but they typically raise usability challenges which, if not
dealt with, may outweigh the intended benefits
Research on semantic web technologies has so far focused mostly on the
technology, but the past 3 years have seen an increase in interest in
interaction design and evaluation
The main part of this talk discusses three of the key usability challenges,
approaches that have been taken to them, and issues that remain open
In the ideal case, users could obtain information via semantic methods by
straightforwardly characterizing their information need in terms of elements of
the ontology(ies) used in the system
In most cases, the ontologies (and other information sources) are too large,
complex, and otherwise unsuited for end-user inspection
Designers of query interfaces for the semantic web have been creative and
often successful in devising ways of allowing users to benefit from the
existence of an ontology without confronting them with its complexity
A goal related to that of reducing effort is the goal of ensuring adequate
expected benefit, which can be relatively difficult with semantically based
interfaces; two strategies are discussed briefly
The same design solutions that reduce effort and complexity in querying can
also make the resulting behavior of the system difficult to understand and
predict, as is illustrated here by a discussion of an intermediate SmartWeb
prototype
Research and experience in human-computer interaction on mental models
yields a number of results and ideas about when and why it is important for a
user to have at least some vague understanding of how a system works internally
and about ways of conveying an appropriate mental model
Many semantic web application scenarios presuppose that some users will invest
effort in providing or enhancing content (e.g., by annotating web pages)
Theory and experience from several fields have yielded a number of ideas about
the conditions under which users may be motivated to
do such work
In more mature fields that involve novel forms of human-computer interaction,
it is often hard to publish a paper concerning a new interactive system unless
it includes some empirical evidence that the novel aspects of the system are
well accepted by users
By contrast, empirical research with users is found only sporadically (though
increasingly) in the semantic web field
Several apparent reasons are discussed
Finally, some general hints about how to involve users effectively in research
on semantic web technologies are given, with emphasis on the diversity of roles
that users can play and the proven effectiveness of interdisciplinary teams in
designing useful and usable systems
|
|
BibTeX entry:
@incollection{Jameson06ESWC,
year = {2006},
author = {{Jameson}, Anthony},
editor = {{Sure}, York and {Domingue}, John B.},
title = {Usability and the Semantic Web},
booktitle = {{The Semantic Web: Research and Applications:
Proceedings of the Third European Semantic Web Conference,
ESWC 2006}},
address = {Berlin},
publisher = {Springer},
pages = {3},
note = {Abstract of an invited talk},
note = {Available from http://dfki.de/$\sim$jameson/abs/Jameson06ESWC.html}}