Case Based Reasoning as a Model for Cognitive Artificial Intelligence

Abstract

Cognitive Systems understand the world through learning and experience. Case Based Reasoning (CBR) systems naturally capture knowledge as experiences in memory and they are able to learn new experiences to retain in their memory. CBR's retrieve and reuse reasoning is also knowledge-rich because of its nearest neighbour retrieval and analogy-based adaptation of retrieved solutions. CBR is particularly suited to domains where there is no well-defined theory, because they have a memory of experiences of what happened, rather than why/how it happened. CBR's assumption that 'similar problems have similar solutions' enables it to understand the contexts for its experiences and the 'bigger picture' from clusters of cases, but also where its similarity assumption is challenged. Here we explore cognition and meta-cognition for CBR through self-reflection and introspection of both memory and retrieve and reuse reasoning. Our idea is to embed and exploit cognitive functionality such as insight, intuition and curiosity within CBR to drive robust, and even explainable, intelligence that will achieve problemsolving in challenging, complex, dynamic domains.

Thinking Fast and Slow: A CBR Perspective

“…Let us examine a few more interesting directions of realizing the slow-fast dichotomy in CBR. As observed earlier, fast thinking is successful in relatively well aligned regions of the case base (Craw and Aamodt 2018), explained in Section 1. This also paves the way for learning better indices The need for realizing the slow-fast dichotomy has been realized outside the CBR context as well, though some of the past work do not explicitly refer to Kahneman's work.…”

“…Problem solving is fast when CBR system responds to a query from regions of high case base alignment. A CBR system slows down when it attempts to solve a query from regions of low case base alignment where retrieved nearest neighbours don't have similar solutions, and adaptation is complex and computationally demanding (Craw and Aamodt 2018). In this paper, we present several novel ways of operationalizing the fast-slow thinking dichotomy by using different decision-making models for each paradigm.…”

Autonomous IoT Device Management Systems: Structured Review and Generalized Cognitive Model

“…To solve a problem, a reasoning mechanism typically uses the principle of analogy to retrieve and reuse episodes that match the current situation. Since case-based reasoning considers what happened rather than on how or why it happened, it is well suited for domains where the context is not explicitly defined [61]. This sets it apart from mechanisms based on semantics.…”

Revisiting Fast and Slow Thinking in Case-Based Reasoning

“…To the best of our knowledge, Craw and Aamodt [4] make the first effort to build a bridge between the dichotomous cognitive mechanisms in Kahneman's work and CBR. The central observation in [4] is that fast thinking can be realized in settings where the knowledge of similarity is straightforward and similar problems are likely to have similar solutions.…”

“…To the best of our knowledge, Craw and Aamodt [4] make the first effort to build a bridge between the dichotomous cognitive mechanisms in Kahneman's work and CBR. The central observation in [4] is that fast thinking can be realized in settings where the knowledge of similarity is straightforward and similar problems are likely to have similar solutions. In contrast, slow thinking may be necessitated when retrieved neighbours have conflicting solutions, and hence more complicated retrieval mechanisms or adaptation knowledge may be involved.…”