Case-based reasoning

Más información sobre el libro

Recording billions of experiences in a database is straightforward, but true wisdom lies in effectively utilizing those experiences. Case-based reasoning (CBR) focuses on experience mining, employing analogical reasoning for problem-solution pairs. Since cases are rarely identical, mere storage and recall are inadequate; thus, defining and analyzing similarity and adaptation is crucial. The fundamentals of CBR are well-established, with numerous successful commercial applications attracting interest from various research fields. This textbook systematically presents CBR with two key goals: to provide rigorous structures for precise reasoning and to showcase a variety of techniques, methods, and tools for diverse applications. Part I introduces the basic elements of CBR without requiring prior knowledge. Part II delves into core methods, including case representations, similarity, retrieval, adaptation, and evaluation. Part III explores advanced topics, addressing uncertainty and probabilities. Part IV discusses various knowledge sources, covering textual CBR, images, sensor data, conversational CBR, and knowledge management. The book concludes with appendices detailing formal definitions and comparisons with other techniques. Drawing on extensive teaching and training experience, the authors include chapter summaries, background notes, and exercises, making it suitable for advanced undergraduate and graduate students in comput