In the Knowledge Services Area, we are researching intelligent methods and services for improving productivity of individual and cooperative knowledge work. We are enhancing personalized information and communication environments with context sensitive recommender services, web 2.0 and collective intelligence. In doing so, we follow a hybrid approach which combines the advantages of semantic technologies with the advantages of heuristic and probabilistic methods. In addition, our methods continuously learn by incorporating implicit usage data and explicit user feedback.

Main Focus:

Our vision of the knowledge management system of the future balances the needs of individual knowledge workers for flexibility with the needs of communities for interconnectivity and the needs of organizations for consistency and efficiency. Ultimately our intelligent knowledge services will support (and incrementally merge) two roles within organizations and communities:

• The individual knowledge worker needs support for executing her current task efficiently and to advance her competencies over time.
• The knowledge engineer needs support for indentifying, modelling and homogenizing knowledge patterns and structures in communities and organizations (e.g. domain ontologies, task structures).

Our service framework KnowSe forms the basis for dynamically orchestrating a large variety of intelligent knowledge services, highly contextualized to a persons work context and interconnected with a multitude of knowledge sources. These knowledge services will operate in a dynamic and ill-defined environment, and will still be able to detect patterns and make inferences based on uncertain information. We will draw on the results of the Area Knowledge Relationship Discovery to realize this ambitious research agenda.

At the moment the following four types of contextualized knowledge services form the core of our research activities:

1. Context Detection Services automatically detect and deduce a person’s work contexts from her user interaction on the desktop. In addition, we will identify several user characteristics, such as goals, interests or prior knowledge from current and past interactions.
2. Contextualized Recommender Services will suggest information resources and persons relevant to the work context of a user and adapted to the user’s characteristics.
3. Contextualized Learning Services provide assistance in informal learning settings in which personal learning or information needs are triggered directly at the workplace.
4. Services for identifying knowledge patterns or structures discover and homogenize emergent knowledge or interaction patterns in community or organizational settings, like domain models or task structures.

Our intelligent knowledge services reinforce the seamless integration of individual and organizational learning, and provide for a grassroots approach to knowledge management.

Scientific Fields of Excellence

• Automatically Detecting a User's Multifaceted Context
• Improving Knowledge Worker Productivity
• Exploiting Emergent Knowledge Structures

Competencies

• Web 2.0, collective intelligence and social network analysis
• Semantic technologies, semantic web and semantic retrieval
• Probabilistic approaches for learning and information needs analysis and user profiling
• Communication and collaboration technologies for knowledge work
• Service-oriented architectures, web services and dynamic orchestration
• Interactive, collaborative modelling environments for ontologies