Abstract
Decision support systems (DSSs) can play a powerful role in natural resource management (NRM), by allowing more effective and collective use of information in addressing complex and often poorly structured questions. CSIRO Tropical Agriculture, in Townsville, Australia, is developing a DSS generator that provides a flexible environment for the construction of decision support tools that assist in assessing the implications of proposed policies or management actions. In this context, the POSEIDON system was designed to help resource managers with the problem formulation phase of building a DSS, i.e. identifying specific issues that need to be analysed to provide an answer to a broad NRM question. This is achieved by a form of knowledge acquisition from free text, which performs intelligent analysis of NRM documents. This article describes the design of POSEIDON and its application in NRM problem formulation.
