Accurate and readily available information is a crucial basis for decision making, problem solving, or performing knowledge-intensive
work. In networked organizations with geographically distributed work force and processes, like logistics networks or SME-clusters,
quickly finding the right information for a given purpose often is a challenge. An improved information supply would contribute
significantly to saving time and most likely to improving productivity. The paper aims at contributing to improved information logistics
by bringing together experiences from knowledge modeling and pattern-based reuse in information system development. We propose a
pattern-based knowledge architecture with several inter-working layers of services for implementing information logistics in networked
organizations. The knowledge architecture forms a framework for selecting and configuring suitable resources for a given problem
situation. The knowledge architecture principle and three types of knowledge patterns within the architecture framework are discussed:
task patterns for representing enterprise knowledge of member organizations in a networked organization, information demand patterns
addressing the information demand of typical roles in a networked organization, and ontology design patterns for capturing context
information for decision support.