Search engines have transformed and improved my work, and with the
notorious unreliability of information on the Web, recommender
technologies have quickly taken hold and promise to be an area of
tremendous growth and importance. Interfaces of the future will no
doubt be far more varied, subtle, and complex than those available
today, and this special issue should contribute to getting there.

Jonathan Grudin, Editor, ACM Transactions on Computer-Human Interaction


Deadline for receipt of manuscripts: 15 July 2003

Recommender Systems Interfaces: Theory and Practice

Papers are invited for a special issue of ACM Transactions on Computer
Human Interaction on the topic of the ways in which Recommender
Systems are changing Human Computer Interaction.  This call is broad in
scope, including HCI research on the psychological and economic theories
of how recommender systems work, interfaces for recommender systems,
experience with recommender systems, and the social issues their
widespread use raises.

The past decade has seen an explosion of interest in recommender
systems, under a wide variety of names including collaborative
filtering, social navigation, personalization, and user modelling. These
systems all share in common the creation of interfaces that adapt
themselves to user behavior based on past experience and patterns of
use.  At their best, recommender systems can create interfaces that
provide each user with the best personalized view of the information,
data, and resources available through the interface. Recommender systems
have been used widely in creating e-commerce sites that present each
user with personalized recommendations, in information resources that
adapt the content they display to each user's tastes, and in navigation
systems that use experience with previous visitors to suggest paths to a
new user.

There are many HCI research challenges in recommender systems, ranging
from interface design to practical experience in e-commerce. Submissions
for this special issue are solicited across the entire range of
recommender systems, including (but not limited to):

1. Social issues in recommender systems.  How does the widespread
    collection of interface usage data affect user privacy?  How do
    users feel about the use of behavioral data to change the interface
    they see?  What can be done to create recommender systems that
    limit the loss of privacy for users?  How does reputation interact
    with recommender systems?

2. Psychological, social, and economic theories about recommender system
    interfaces.  What does theory tell us about the usage patterns for
    recommender systems?  Can we predict which interfaces will be
    successful and which will fail?  What new theory is needed to
    explain the experiences with recommender systems to date?

3. Design of interfaces for recommender systems.  What are best
    practices for recommender system design for different user tasks?
    Should recommender systems be visible to users, or silent servants
    behind the scene?

4. Experience with recommender systems.  How effective are recommender
    systems in practice?  What changes do they create in usage
    patterns?  In which domains do they work well?  What limits are
    there on their utility?


TOCHI's normal rigorous journal refereeing standards will apply. Though
the intent is to provide broad coverage over the area of recommender
systems, only papers that meet TOCHI's normal standards will be
included.

Submission format: TOCHI uses digital submission and distribution of
manuscripts. Details of the submission process can be found on the TOCHI
website at http://www.acm.org/tochi/. ACM requires digital submission of
accepted papers in one of several formats. These formats are specified
at the ToCHI website. To facilitate the publication process, authors are
advised to use one of these formats when submitting their manuscript. Be
sure to indicate in your cover letter that you are submitting your
manuscript for the special issue on Recommender Systems.

If you have any questions about the special issue, please contact the
editors:

John Riedl        Paul Dourish
[log in to unmask]  [log in to unmask]

Deadline for receipt of manuscripts: 15 July 2003.