Information compression by multiple alignment, unification and search as a framework for human-like reasoning
Guntram B. Wolff · Logic Journal of IGPL · 2001
This paper is an overview of the idea that several aspects of human-like reasoning may be understood in a framework of information compression by 'multiple alignment', 'unification' and 'search'. These concepts will be described in outline together with a description of the SP61 computer model which is a partial realisation of the framework. A selection of examples are presented, illustrated with output from the SP61 model.