This paper firstly provides a re-appraisal of the development of techniques for inverting deduction, secondly introduces Mode-Directed Inverse Entailment (MDIE) as a generalisation and enhancement of previous approaches and thirdly describes an implementation of MDIE in the Progol system. Progol is implemented in C and available by anonymous ftp. The re-assessment of previous techniques in terms of inverse entailment leads to new results for learning from positive data and inverting implication between pairs of clauses.

This paper is a survey of inductive rule learning algorithms that use a separate-and-conquer strategy. This strategy can be traced back to the AQ learning system and still enjoys popularity as can be seen from its frequent use in inductive logic programming systems. We will put this wide variety of algorithms into a single framework and analyze them along three different dimensions, namely their search, language and overfitting avoidance biases.

OPUS is a branch and bound search algorithm that enables efficient admissible search through spaces for which the order of search operator application is not significant. The algorithm's search efficiency is demonstrated with respect to very large machine learning search spaces. The use of admissible search is of potential value to the machine learning community as it means that the exact learning biases to be employed for complex learning tasks can be precisely specified and manipulated. OPUS also has potential for application in other areas of artificial intelligence, notably, truth maintenance.

Program specialisation aims at improving the overall performance of programs by performing source to source transformations. A common approach within functional and logic programming, known respectively as partial evaluation and partial deduction, is to exploit partial knowledge about the input. It is achieved through a well-automated application of parts of the Burstall-Darlington unfold/fold transformation framework. The main challenge in developing systems is to design automatic control that ensures correctness, efficiency, and termination. This survey and tutorial presents the main developments in controlling partial deduction over the past 10 years and analyses their respective merits and shortcomings. It ends with an assessment of current achievements and sketches some remaining research challenges.

This paper describes an algorithm performing an analysis and transformation of logic programs. The transformation achieves two goals: redundant functors are removed from the program, and procedures may be split into two or more specialised versions handling different cases. It can be applied to most logic programming languages, including concurrent logic programming languages, because the transformations perform no unfolding of the program; they only remove some redundant operations within the unifications. The main saving is in heap usage, though time performance may also be improved. One of the main purposes of the transformation is to “clean up ” programs generated by other methods of transformation or synthesis. The analysis is an example of an abstract interpretation, and is guaranteed to terminate. A Prolog implementation of the algorithm, illustrating some meta-programming techniques, is given and some results are reported.

In this paper we address the problem of efficiently implementing HiLog, a logic programming language with higher-order syntax and first-order semantics. In contrast to approaches proposed in the literature that modify, or abandon the WAM framework in order to implement HiLog, our approach to the problem stems from a belief that the WAM should be an adequate abstract machine for the execution of any logic language with first-order semantics. To show how to implement HiLog by staying within the WAM framework, we identify the reasons for poor performance characteristics of HiLog programs, present requirements for efficient HiLog execution, and propose a complete solution to the problem. Our proposal, which can be viewed either as a compile-time program specialisation preprocessing step, or as an enhancement to the HiLog encoding in predicate calculus presented by Chen, Kifer, and Warren in [1], allows HiLog to be efficiently implemented on any Prolog system by simply modifying Prolog's in...

This paper describes a new algorithm for learning decision lists that operates by prepending successive rules to front of the list under construction. This contrasts with the original decision list induction algorithm which operates by appending successive rules to end of the list under construction. The new algorithm is demonstrated in the majority of cases to produce smaller classifiers that provide improved predictive accuracy than those produced by the original decision list induction algorithm.

The general claims of this paper are twofold: there are challenging problems for Machine Learning in the field of Databases, and the study of these problems leads to a deeper understanding of Machine Learning. To support the first claim, we consider the problem of characterising a database relation in terms of high-level properties, i.e. attribute dependencies. The problem is reformulated to reveal its inductive nature. To support the second claim, we show that the problems presented here do not fit well into the current framework for inductive learning, and we discuss the outline of a more general theory of inductive learning. KEYWORDS: Relational model, attribute dependencies, inductive learning, theoretical analysis. Contents 1. Introduction.......................................................................................................................1 2. Characterising a database relation..........................................................................................

Relational learning algorithms are of special interest to members of the machine learning community; they offer practical methods for extending the representations used in algorithms that solve supervised learning tasks. Five approaches are currently being explored to address issues involved with using relational representations. This paper surveys algorithms embodying these approaches, summarizes their empirical evaluations, highlights their commonalities, and suggests potential directions for future research. Keywords: supervised learning, representation, relational learning 1 Introduction Relational learning algorithms extend the capabilities of propositional or monadic supervised learning algorithms. Supervised learning algorithms input a set of instances, which are described by a set of predictor descriptors and a target descriptor. These algorithms construct a function (i.e., a concept description) that can predict an instance's target descriptor value given its predictor desc...

One central point of machine learning in general and inductive logic programming in special is the search space of the algorithms, defined by the control structure of the algorithms and additional knowledge. Since the sensible search space differs from domain to domain, a flexible way to describe this space is desired. To demonstrate problems occuring while using existing algorithms, we introduce learning tasks in a real world domain: concept learning for navigation of autonomous mobile robots. We point out differences between three existing algorithms used within this framework and their results. Since all of these algorithms have problems in solving the tasks, we developed grdt (grammar based rule discovery tool), an algorithm combining their ideas and techniques. In grdt a two level description language is used for describing the hypothesis space. A grammar is used to define a set of second order rule schemata and these schemata then define the hypothesis space itself. 1 Introductio...