Kolflow

From Kolflow Project

Revision as of 17:40, 27 January 2011 by Desmontils-e (Talk | contribs)
Jump to: navigation, search

Web 2.0 is currently producing a huge amount of information. Continuously transforming this information into knowledge is a major challenge for the research community. Automated reasoning or collective intelligence are currently two representative approaches to transform content into knowledge. The Kolflow project proposes to extend collective intelligence with smart agents relying on automated reasoning. Smart agents can significantly reduce the overhead of communities in the process of continuously building knowledge. Consequently, continuous knowledge building is much more efficient.

Kolflow aims at building a social semantic space where humans collaborate with smart agents in order to produce knowledge understandable by humans and machines. Humans are able to understand the actions of smart agents. Smart agents are able to understand actions of humans. Kolflow targets the co-evolution of content and knowledge as the result of interactions of humans and machines.

If human-machine collaboration can be the key to ensure co-evolution of content and knowl- edge, such collaboration can fail if not managed. The Kolflow project addresses the following scientific issues:

Man-machine collaboration
Man-machine collaboration can be very unstable and make the whole system divergent. How to coordinate the actions of distributed agents, either software or humans, sharing web contents and knowledge accessed by web users at a potentially large scale? In particular, a key issue is to guarantee a minimal stability and the non-regression of the whole system.
Man-machine collaboration for humans
how to make formal knowledge and its evolution accessible, usable, editable and understandable by human agents so they can observe, control, evaluate and reuse the outputs of smart agents?
Man-machine collaboration for machines
how to support and take into account the unpre- dictable behavior of human agents that can at any moment add or modify content and formal knowledge with the risk of introducing uncertainty or inconsistency? How auto- mated reasonings can adapt their behavior and results by taking into account feedback from human agents? How these tools can adapt their behavior and results to specific user needs in a given context?

The Kolflow project aims at tackling man-machine collaboration issues with the following approach:

  1. Deliver man-machine collaboration scenarios and some reference corpus. These scenarios drive the project and evaluate the overall progression of others Kolflow tasks.
  2. Build a social semantic space based on distributed semantic wikis. This social semantic space behaves as a blackboard for man-machine collaboration. Coordination of agents is based on process representation and enactment. The whole system is accessible by humans and machines through distributed semantic queries.
  3. Make histories of knowledge building understandable by man and machine. This is the key to make smart agent aware of humans reactions to their actions.
  4. Make automated reasoning understandable by humans. Smart agents must explain what they did and why they did these actions to humans.
  5. Manage inconsistencies generated by man-machine collaboration by allowing interactive reasoning with a globally inconsistent family of ontologies.

Expected scientific results of the Kolflow project will be a number of publications in the area of semantic web, computer-supported cooperative work, and knowledge discovery and data mining. The technical result of the project will be fundamental and practical knowledge in man-machine collaboration. It includes a basic corpus and scenarios and prototypes that enacts man-machine collaboration scenarios. If successful, Kolflow will demonstrate how it is possible to extend collective intelligence with smart agents in order to ensure co-evolution of contents and knowledge.