DataTerminology.comNEW
Also known as: CBR
A problem-solving method that reuses solutions from similar past cases to address new problems.
DESCRIPTION
Case-based reasoning (CBR) solves new problems by retrieving and adapting solutions from similar prior cases. Typical CBR cycles include retrieve, reuse, revise, and retain. It is applied in help desks, medical diagnosis, legal reasoning, and recommender systems. CBR complements rule-based or model-driven methods when explicit rules are incomplete or costly to encode. Effective CBR requires robust similarity metrics, case indexing, and case-base maintenance.
ALTERNATIVE DEFINITIONS
Wikipedia
Case-based reasoning (CBR), broadly construed, is the process of solving new problems based on the solutions of similar past problems.
Emphasizes reuse of solutions from similar prior cases as the core mechanism.
Aamodt & Plaza (1994), AI Communications
Case-based reasoning is a problem solving paradigm that in many respects is fundamentally different from other major AI approaches.
Highlights CBR as a distinct paradigm in AI, contrasting it with rule-based and model-based approaches.
RELATED TERMS
K-Nearest Neighbors Similarity Measure Analogy-Based Reasoning Knowledge Base Retrieval-Augmented Generation Problem Solving Reasoning by Analogy Instance-Based Learning Fine-Tuning
COMPETING TERMS
Rule-based Reasoning Model-Based Reasoning
RELATIONS
- Broader: Reasoning
KNOWLEDGE AREAS
TARGET AUDIENCE
AI Engineer Data Scientist MLOps Engineer AI Product Manager
EXTERNAL SOURCES
- Wikipedia - Case-based reasoning
- Aamodt & Plaza - Foundational CBR survey paper
- Springer Encyclopedia - Reference entry on CBR
- scikit-learn - k-NN instance-based learning docs
TOP QUESTIONS
Why is Case-Based Reasoning still relevant for AI systems?
CBR excels when rules are incomplete and data is sparse. It leverages organizational experience, reduces modeling effort, and supports explainability through case precedents.
Category: Relevancy
How is a Case-Based Reasoning system typically structured?
It follows the retrieve–reuse–revise–retain cycle. Cases are indexed, similar cases are retrieved, adapted to new problems, evaluated, and then retained for future use.
Category: Management
What data is required to implement Case-Based Reasoning?
A well-curated case base with problem descriptors, solutions, and outcomes is essential. You also need similarity metrics and adaptation rules or strategies.
Category: Requirements
What are common mistakes when applying Case-Based Reasoning?
Poor case indexing and weak similarity measures lead to irrelevant retrievals. Failing to maintain the case base causes drift and degraded performance.
Category: Mistakes
What risks accompany Case-Based Reasoning in production?
Bias in historical cases can propagate unfair decisions. Without governance and monitoring, outdated cases misguide future recommendations.
Category: Risks
How does Case-Based Reasoning differ from rule-based systems?
CBR relies on past cases and similarity, whereas rule-based systems depend on explicit, often brittle, expert rules. CBR adapts more flexibly to novel problems.
Category: Differences
How can Case-Based Reasoning integrate with Retrieval-Augmented Generation?
CBR can supply curated, structured cases as retrievals, while a generator composes responses. This combination enhances relevance and explanations.
Category: Application
Where should I start learning about Case-Based Reasoning?
Begin with Aamodt and Plaza’s survey and explore example implementations like k-NN baselines and domain-specific case libraries.
Category: Getting Started
What is a striking analogy for Case-Based Reasoning?
It’s like consulting a well-organized notebook of past fixes to repair a new device, then updating the notebook with what worked.
Category: Example