Computer learning in theorem proving

Trigonometric theorem proofs are taken as the application for the learning model described in this paper. The simple basic structure of the heuristics together with the detailed implementation are developed, with evaluation as to possible generality in more general problem-solving processes.