This paper provides an overview of the development of Knowledge Engineering (KE) over the last 15 years. It discusses the shift from a transfer approach to a modeling approach and describes two key developments: Role-limiting Methods and Generic Tasks. The paper also presents three modeling frameworks—CommonKADS, MIKE, and PROTÉGÉ-II—and discusses important methodological developments such as specification languages, problem-solving methods, and ontologies. It concludes with an outline of the relationship between KE and other disciplines like Software Engineering, Information Integration, and Knowledge Management. The paper begins by discussing the historical roots of KE, emphasizing the shift from a transfer approach to a modeling approach. The transfer approach, which focused on transferring human knowledge into an implemented knowledge base, was found to be inadequate for large, reliable knowledge bases. This led to the development of the modeling approach, which views the construction of a KBS as a modeling activity. This approach emphasizes the need for a model that can be refined and revised based on feedback from the real world. The paper then discusses the concept of Problem-Solving Methods (PSMs), which are generic inference patterns that describe the problem-solving behavior of expert systems. It describes the PSM Heuristic Classification and its application in different domains. The paper also discusses the concept of Role-Limiting Methods (RLMs), which are based on PSMs and allow for the reuse of knowledge in different domains. It also discusses the concept of Generic Tasks (GTs), which are building blocks that can be reused for the construction of different KBSs. The paper then presents three modeling frameworks: CommonKADS, MIKE, and PROTÉGÉ-II. CommonKADS is prominent for defining the structure of the Expertise Model, MIKE emphasizes the formal and executable specification of the Expertise Model, and PROTÉGÉ-II exploits the notion of ontologies. The paper also discusses the importance of ontologies in KE and their role in supporting knowledge acquisition. The paper concludes by discussing the importance of specification languages in KE, which allow for the precise description of knowledge and reasoning processes. It also discusses the need for formal specification techniques to overcome the limitations of natural language descriptions. The paper highlights the importance of combining non-functional and functional specification techniques in KE to ensure the effective and efficient implementation of KBSs.This paper provides an overview of the development of Knowledge Engineering (KE) over the last 15 years. It discusses the shift from a transfer approach to a modeling approach and describes two key developments: Role-limiting Methods and Generic Tasks. The paper also presents three modeling frameworks—CommonKADS, MIKE, and PROTÉGÉ-II—and discusses important methodological developments such as specification languages, problem-solving methods, and ontologies. It concludes with an outline of the relationship between KE and other disciplines like Software Engineering, Information Integration, and Knowledge Management. The paper begins by discussing the historical roots of KE, emphasizing the shift from a transfer approach to a modeling approach. The transfer approach, which focused on transferring human knowledge into an implemented knowledge base, was found to be inadequate for large, reliable knowledge bases. This led to the development of the modeling approach, which views the construction of a KBS as a modeling activity. This approach emphasizes the need for a model that can be refined and revised based on feedback from the real world. The paper then discusses the concept of Problem-Solving Methods (PSMs), which are generic inference patterns that describe the problem-solving behavior of expert systems. It describes the PSM Heuristic Classification and its application in different domains. The paper also discusses the concept of Role-Limiting Methods (RLMs), which are based on PSMs and allow for the reuse of knowledge in different domains. It also discusses the concept of Generic Tasks (GTs), which are building blocks that can be reused for the construction of different KBSs. The paper then presents three modeling frameworks: CommonKADS, MIKE, and PROTÉGÉ-II. CommonKADS is prominent for defining the structure of the Expertise Model, MIKE emphasizes the formal and executable specification of the Expertise Model, and PROTÉGÉ-II exploits the notion of ontologies. The paper also discusses the importance of ontologies in KE and their role in supporting knowledge acquisition. The paper concludes by discussing the importance of specification languages in KE, which allow for the precise description of knowledge and reasoning processes. It also discusses the need for formal specification techniques to overcome the limitations of natural language descriptions. The paper highlights the importance of combining non-functional and functional specification techniques in KE to ensure the effective and efficient implementation of KBSs.