Reasoning with positive and negative cases (2020)
Supervisors: Emmanuel Nauer and Jean Lieber (email@example.com), team K
Keywords: case-based reasoning, positive and negative cases, knowledge representation and reasoning, symbolic machine learning
Summary of the subject:
Case-based reasoning (CBR) consists in solving problems using cases, a case being the representation of a problem-solving episode in a given domain. Classically, it is assumed that cases from the case base are positive in the sense that they constitute an acceptable solution for the user. The objective of this thesis is to examine how the CBR principle can evolve in order to take into account positive and negative cases. An application will be developed to validate this work in one of the application domains of team K.