Report of the Breakout Group 
                            on 
"Recommendations for IDM-Related Research Initiatives"

Group Coordinators: David D. Lewis and Ouri Wolfson

This discussion group considered new and modified emphases that might be
appropriate in IDM's research directions.  To a surprising extent, the
directions proposed were as much in reaction to developments in technology,
as anticipatory of them.

Traditionally, work at the foundations of computer science opened up new
technological possibilities, which were eventually transferred and deployed
commercially.   This model has been joined by one where the commercial
world, particularly in entrepreneurial areas like the Internet and wireless,
is apt to run past the research community into immediate deployment of new
models of information management and access.  The danger, of course, is that
expensive, awkward, and unsustainable approaches will become the facts on
the ground before an understanding of their foundations, of the range of
possible approaches, is understood. Thus basic research in these areas is
more critical than ever.

The following areas were proposed, as ones where past progress has set the
stage for new advances, and where rapidly moving technology has made those
advances particularly critical:


Focus areas:

1. Interactivity: visualization for  naive users, exploration and
discovery, paradigms such as travel through data and IT-as-tour-guide,
personalization and heterogeneity.

2. Anytime/anywhere/anyone access to data: accessibility, wireless access
by mobile users, triggers and push technology,
multilingual users.

3. Intelligent resource usage and resource aware data management: 
resources include bandwidth, storage, display area; techniques
include compression and query optimization.

4. Data mining: heterogeneous, semistructured, one pass, dynamic,
distributed, interactive, data mining;
grounding of text data (i.e. connecting textual
entities with real world ones).

5. Context-aware data management:
context includes geography, etc.; databases containing information on
dynamically changing collections of entities (automobile traffic, network
connectivity, etc.), communities - lots of dynamically interacting entities
in the data; data management in infrastructureless wireless environments,
i.e. processing nodes communicating wirelessly, without a 
cellular infrastructure.

6. Summarization: data reduction, 
personalization - who are you summarizing for, summarization in context, 
the summarization pyramid as a grand notion.

7. Other important issues: metadata, security/privacy/access authorization,
authentication, confidence and data reliability, dynamic schema discovery
and extraction, trust, IR and multimedia