Our research interests in this project are in exploring how automated reasoning systems can learn theorem proving strategies. In particular, we are looking into how a proof planning system (Bundy, 1988) can automatically learn

We present the NumbersWithNames program which performs data-mining on the Encyclopedia of Integer Sequences to nd interesting conjectures in number theory. The program forms conjectures by nding empirical relationships between a sequence chosen by the user and those in the Encyclopedia. Furthermore, it transforms the chosen sequence into another set of sequences about which conjectures can also be formed. Finally, the program prunes and sorts the conjectures so that the most plausible ones are presented rst. We describe here the many improvements to the previous Prolog implementation which have enabled us to provide NumbersWithNames as an online program. We also present some new results from using NumbersWithNames, including details of an

In this paper we present a framework for automated learning within mathematical reasoning systems. In particular, this framework enables proof planning systems to automatically learn new proof methods from well chosen examples of proofs which use a similar reasoning strategy to prove related theorems. Our framework consists of a representation formalism for methods and a machine learning technique which can learn methods using this representation formalism. Our aim is to emulate some of the human informal mathematical reasoning, in particular the human learning capability on machines. This work bridges two areas of research, namely it applies machine learning techniques to advance the capability of automated reasoning systems. 1 Introduction Proof planning [4] is an approach to theorem proving which uses proof methods rather than low level logical inference rules to nd a proof of a theorem at hand. A proof method species and encodes a general reasoning strategy that can be ...