| Virtual Personal Assistants - Siri |
The Siri story has just begun. |
| Social Knowledge
Sharing - RealTravel |
RealTravel is an example of
domain specific knowledge sharing in the consumer Internet
space. RealTravel.com aspires
to be the best place on the web to share knowledge and experiences
about travel. The site provides an environment for a community
of travel enthusiasts to create beautiful travel journals
of their adventures, share them with friends and family,
and find other like-minded travelers. People looking for
information about where to go, where to stay, or what to
do in their travel can learn from the authentic experiences
of those who have been there.
The site applies technologies from asynchronous collaboration,
knowledge management, social networks, GIS and mapping, tagging,
content management, and machine learning. For more information, see
the RealTravel Story. |
| Collaborative
Knowledge Management - Intraspect |
Collaborative Knowledge
Management, a pioneering approach to enabling organizational
intelligence in the enterprise, is now sold as commercial
products in several public companies. The key insight was
harnessing the value of collaborative work to create collective
intelligence. Tom invented the basic idea of CKM, designed
the Intraspect architecture, and worried about the user experience.
Over eight years, about 200 people implemented, refined, and
customized the product for hundreds of thousands of users.
For more information, see the Intraspect
story. |
| Ontologies as technology
and the Knowledge Sharing stack |
In the early 1990's, a group
of AI and DB groups got together to define a
standard architecture stack for allowing intelligent
systems to interoperate over a knowledge bus and share data,
models, and other knowledge without sharing data schema or
formats (today it's called Semantic
Web). Tom was part of
a collaborative team comprised of colleagues from different
institutions and companies, and he co-led the effort for
the part of the stack related to ontology sharing. He gets
cited a lot for clarifying
a
definition of ontology for the AI community. See papers
on ontology
sharing,
ontology design,
and
ontologies as standards.
read more>>
|
| Simple collective memory
- Hypermail |
One of the first integrations of
the web with email, Hypermail turns standard mailing lists
into published web accounts and institutional records of conversations.
It was distributed freely
and widely in the early web years, and is now has an active open
source community extending it. Tom built the first version
in Lisp in 1994, and Kevin
Hughes rewrote it in C and convinced management to release
it as open source. It was used to publish and archive the
early email discussions (such as www-talk) that shaped the
creation of the WWW. Once released to the public, it took
off. read more>>
|
| Virtual
Documents |
Virtual
Documents are a user interaction medium for knowledge
systems that exploits the (then) new capabilities of the
Web for dynamic hypermedia and the AI techniques for model-based
generation of natural language explanations. In contrast
to convention GUIs, the virtual document dynamically generates
its content as a dialog with the reader. The content is
full of hyperlinks, which the user follows to follow up
with questions or move to the next level of detail. Tom
wasn't the first to show a dynamic web site, but he was
a vocal proponent of the use of generative
dynamic interfaces as a communication medium.
See papers 1 ,
2,
3,
4 |
| Knowledge Acquisition |
The main reason knowledge systems
did not become ubiquitous (compared with relational database
systems) is the knowledge acquisition bottleneck: it's hard
to build systems that model and act on knowledge of human
experts. Tom worked on systems to help knowledge systems build
up their knowledge by having them interact with experts in
the context of solving problems. see papers |
| Knowledge-based Communication Prosthesis |
One the worst things about
being severely disabled is the difficulty with communication.
Conventional communication prosthesis systems are either
too slow (essentially one-button keyboards) or too restrictive
(limited to a small, fixed vocabulary). In the early 1980's,
Tom built a system that could use knowledge of the domain
of discourse of a person to help dynamically suggest menu
choices so the one-button input could produce useful output
in a reasonable time. |
 |
|
