Abstract. An intelligent hybrid system is proposed. It includes an adaptive human-machine interface and a hybrid case-based reasoning component for knowledge engineering. The adaptive human-machine interface remembers past question and answer scenarios between the system and the end users. It employs a KA-related commonsense base to help user interaction and a housekeeping engine to search and aggregate the data relevant to those the user answered. It discriminates the professional level of the user by the fuzziness of his answers, and provides different interaction patterns for the user accordingly. The hybrid case-based reasoning component hybridizes case-based reasoning, neural networks, fuzzy theory, induction, and knowledge-based reasoning technology. Hybridizing these techniques together properly enhances the robustness of the system, improves the knowledge engineering process, and promotes the quality of the developed knowledge-based systems. L IntroductionKnowledge acquisition (KA) has long been recognized as a very hard problem in knowledge engineering. Many techniques have been proposed, e.g., GAS [3], PROTEGE II [14] and KADS [17] to alleviate it. Most of them are meta-tools. The responsibility of the knowledge engineer is to select appropriate building blocks including problem solving methods, KA methods, knowledge representation methods, and inference methods. It is so easy for a junior knowledge engineer to be trapped in his stereotyped behavior during the selection of the building blocks that he is unable to fully exploit the power of the meta-tools. In general, the quality of the developed KA-tools using these meta-tools is heavily influenced by his experience.It looks a good idea that a technique that can accumulate knowledge engineering experience is introduced to help the knowledge engineers in using a meta-tool. Case-based reasoning is such a method. It relies on previous experience to solve a problem [7,10]; it comes up with new solutions by adapting the old solutions that have successfully solved previous similar problems. However, a knowledge engineering scenario usually involves much information. Which of it is significant enough in recognizing the previous cases for use is a big problem. How to maintain a case library so that it can be efficiently and effectively used poses another problem. In this paper, we propose a hybrid case-based system that integrates neural networks, induction, and knowledge-based technology to help learn how to focus on the problem and how to narrow down likely hypotheses in a way similar to the expert