We believe that situation theory as described by Devlin in his recent book, Logic and Information (1991) may provide a useful approach to ontological engineering. Situations provide a flexible method of knowledge representation well suited to take advantage of a light weight parse of natural language into phrase units. With this approach it is our purpose to design a system that takes advantage of as much automatic processing as possible to aid in knowledge acquisition but which relies ultimately on the human to correct the process. As the knowledge base grows it will in turn be used to aid in the process of acquisition. As conceived the initial stages will be the most labor intensive. The first task for the system will be to test its usefulness on a small set of documents to determine how it may aid retrieval. If results are positive we may proceed to a more ambitious effort to build a large base of information in molecular biology.
