This is a placeholder.

Alice is an open source project for radical improvements in user interaction. Targeted applications are "folded" metamodels, especially the many types of "meta" models identified in the ICEIMT workshops (and followup activity) for complex enterprise modeling. Addresses the specific scenario of semantically-informed self-organizing systems explicated in the ALF area, and leverages similar technologies.

Initial application: B-Man (Mobile Agents for Workflow)

Implementation languages: Haskell (or Erlang) on serverside, Javascript/ECMAscript clients. ARC-like abstraction. Konfabulator-like modularization.

ALICE depends on work in the cognitive sciences, some of which is reported in the FilmsFolded and GeoKabbalitter sections.

University host and consulting company to be announced.

Here will be all the results from the DARPA study on Agile Virtual Enterprises (but no code); all the insights from the Joint US/EC ICEIMT work; all the work of the group toward semantically informed self-organizing systems work. There may be some FIS stuff here as well, and of course, the ALF pages from sourceforge.

As with the FilmsFolded and Geokabbalitter sections, the "live" part of the site will be Tinderbox and Keynote projects.

A substantial amount of interesting information from the previous site is here.

What Problem Does ALICE Address?

What Problem does ALICE Address?

Consider the case of large amounts of information, perhaps from dynamic, distributed, diverse sources. Many times, a user has a means to access and collect this information. We address this situation where:

• The sources and forms of information are determined by at least partially non-deterministic means, for instance with distributed agent systems.
• Some of the information is metainformation, including metainformation about why it may be relevant and how it fits in.
• A client may have a unique set of needs for translating, inter-relating, displaying and browsing that information. And there may be non-deterministic processes at the client side in requesting, evaluating and structuring that information.
• That client would find it useful to have smooth transitions from detailed views to “big picture” views, and from simple views to reflexive or “folded” views that contain both the information but metainformation on why it was related (and displayed) in the context employed.
• Finally, though the client may be a specialist in some discipline, he/she is not an expert modeler/computer user.

One user example may be a client computer for a decisionmaker in an advanced virtual enterprise where lots of small businesses have agents that act selfishly to try combinations of thousands of potential partners, thousands (or millions) of product/service configurations to discover and optimize opportunities. Some of this may be through guidance from humans, each on their clients. Each client may have completely unique ontologies, modeling methods, business practices and cultural influences. Except for some modicum of semantic standard, few limiting standards are mandated, this as a deliberate decision to gain competitive advantage and get the largest possible pool.

ALICE might usefully be on some of these clients, receiving information from agent systems (See ALF and B-Man), structuring it on the fly using local meta-agents, a locally active geometric cosmology tied to a PSL-mediated formal ontology and individually useful metaphoric concept indices..

Intelligence Community uses are easily envisioned.