Abstract:Induction in saturation-based first-order theorem proving is a new exciting direction in the automation of inductive reasoning. In this paper we survey our work on integrating induction directly into the saturation-based proof search framework of first-order theorem proving. We describe our induction inference rules proving properties with inductively defined datatypes and integers. We also present additional reasoning heuristics for strengthening inductive reasoning, as well as for using induction hypotheses and recursive function definitions for guiding induction. We present exhaustive experimental results demonstrating the practical impact of our approach as implemented within Vampire.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to directly integrate inductive reasoning into the saturation - based first - order theorem - proving framework to automate the proof of properties of inductively defined data types (such as natural numbers) and integers. Specifically, the authors aim to:
1. **Integrate inductive reasoning into the saturation proof - search framework**: Previous work has mainly focused on when to apply induction and which induction axioms to use, while this paper is committed to directly introducing inductive - reasoning rules in the saturation - proof process.
2. **Develop new inductive - reasoning rules**: In order to enhance the ability of inductive reasoning, the authors propose inductive - reasoning rules for inductively defined data types and integers, and combine recursive - function definitions to guide induction.
3. **Optimize the efficiency of inductive reasoning**: By introducing multi - premise inductive - reasoning rules, generalizing induction, and using inductive hypotheses as rewrite rules, etc., the efficiency of saturation - based inductive reasoning is improved.
4. **Experimentally prove the effectiveness of the method**: Through a series of experiments, show the application effect of their method in practical problems, especially those problems that could not be solved by other systems before.
### Core contributions of the paper
- **Combination of saturation and induction**: A new method is proposed to directly integrate inductive reasoning into the saturation - proof process, enabling first - order theorem provers to handle inductive reasoning more effectively.
- **Extension of inductive - reasoning rules**: Multiple inductive - reasoning rules are developed, including multi - premise induction, generalized induction, etc., enhancing the reasoning ability of the system.
- **Efficient reasoning strategies**: By introducing new reasoning rules and optimization techniques, the efficiency of inductive reasoning in saturation - proof is ensured.
### Example analysis
The paper shows the application of its method through a specific example. Consider the following function program:
```plaintext
datatype nat = 0 | s(x)
fun add(0, y) = y
| add(s(z), y) = s(add(z, y))
fun even(0) = ⊤
| even(s(z)) = ¬even(z)
fun half(0) = 0
| half(s(0)) = 0
| half(s(s(z))) = s(half(z))
```
The verification task is to prove that for all natural numbers \(x\), if \(x\) is an even number, then \(x = add(half(x), half(x))\). That is to prove the formula:
\[ \forall x \in \text{nat}. (\text{even}(x) \rightarrow x = \text{add}(\text{half}(x), \text{half}(x))) \]
This requires inductive reasoning on natural numbers. The paper describes in detail how to complete this proof by introducing appropriate induction patterns and reasoning rules.
### Summary
The main goal of this paper is to realize the automation of inductive reasoning by directly integrating it into the saturation - based first - order theorem - proving framework, and to improve its efficiency and application scope through a series of technical means.