The Generic Rough Set Inductive Logic Programming Model and Motifs in Strings

The Variable Precision Rough Set Inductive Logic Programming model (VPRSILP model) extends the Variable Precision Rough Set (VPRS) model to Inductive Logic Programming (ILP). The generic Rough Set Inductive Logic Programming (gRS‐ILP) model provides a framework for ILP when the setting is imprecise and any induced logic program will not be able to distinguish between certain positive and negative examples. The gRS‐ILP model is extended in this paper to the VPRSILP model by including features of the VPRS model. The VPRSILP model is applied to strings and an illustrative experiment on transmembrane domains in amino acid sequences is presented.

The example semantics of Inductive Logic Programming (ILP) systems is said to be in a rough setting when the consistency and completeness criteria cannot both be fulfilled together, because the evidence, background knowledge and declarative bias are such that any induced hypothesis cannot distinguish between some of the positive and negative examples. The gRS-ILP model (generic Rough Set Inductive Logic Programming model) provides a theoretical foundation in this rough setting for an ILP system to induce hypotheses that are used to say that an example is definitely positive, or definitely negative. An illustrative example using Progol is presented. Results are presented of GOLEM experiments using the data set for drug design for Alzheimer's disease and other experiments using Progol on mutagenesis data and transmembrane domain data.

The Variable Precision Rough Set Inductive Logic Programming model (VPRSILP model) is the extension of the Variable Precision Rough Set (VPRS) model to Inductive Logic Programming (ILP). The generic Rough Set Inductive Logic Programming (gRS-ILP) model provides a framework for ILP when the setting is imprecise and any induced logic program will not be able to distinguish between certain positive and negative examples. The gRS-ILP model is extended to the VPRSILP model by including features of the VPRS model. The VPRSILP model is further extended in this paper to include future test cases. This model is applied to Web usage mining and an illustrative experiment is presented.

Inductive Logic Programming: Issues, results and the challenge of Learning Language in Logic

Some inductive logic programming (ILP) systems use determinate literals to efficiently induce logic programs. A determinate literal literal is a literal that does not distinguish positive examples from negative examples, but produces information in variables introduced by the literal. The concept of determinate literals, however, is not reflected by the concept of input/output mode of predicate attributes properly, and so a system using determinate literals may induce inconsistent logic programs with predicate mode or inexecutable programs. The paper extends the concept of determinate literals and proposes input and output determinate literals. These literals function as pre-processor and post-processor against other literals. The paper also describes an implementation of the method and experimentations.

Inductive Logic Programming (ILP) induces first-order clausal theories from learning examples (positive and negative) and knowledge of the domain. Such theories can be used for gradually increasing the understanding of a robot about its world through iterative learning process. In this work we present a method for autonomous creation of negative examples for an ILP learner (i.e. sifting method). We also present a method for managing the iterative learning process (i.e. step transition method) for an autonomous robot that uses the ILP learner for learning. The sifting method uses `out-of-domain' values of the parameters involved in the learning process to create a set of possible negative examples. From these examples the robot autonomously selects those which allow it to efficiently learn better hypotheses. The step transition method enables the autonomous robot to decide how should it learn the new knowledge such that it can also gain profit from its experience. The robot makes this decision by comparing results of two different learning processes conducted on different data sets produced by the same action of the robot. The proposed methods are developed using the ILP learner Aleph, however they can also be used with other similar ILP learners. We experiment with these methods for learning primitive concepts for a mobile autonomous robot in a simple world. Results of the experiments show that the robot learns meaningful definitions of different physical notions in a hierarchical manner.

Automatic traffic incident detection based on nFOIL

Logics and Multi-agents: towards a new symbolic model of cognition

Learning first-order logic programs from relational facts yields intuitive insights into the data. Inductive logic programming (ILP) models are effective in learning first-order logic programs from observed relational data. Symbolic ILP models support rule learning in a data-ecient manner. However, symbolic ILP models are not robust to learn from noisy data. Neuro-symbolic ILP models utilize neural networks to learn logic programs in a differentiable manner which improves the robustness of ILP models. However, most neuro-symbolic methods need a strong language bias to learn logic programs, which reduces the usability and flexibility of ILP models and limits the logic program formats. In addition, most neuro-symbolic ILP methods cannot learn logic programs effectively from both small-size datasets and large-size datasets such as knowledge graphs. In the paper, we introduce a novel differentiable ILP model called differentiable first-order rule learner (DFORL), which is scalable to learn rules from both smaller and larger datasets. Besides, DFORL only needs the number of variables in the learned logic programs as input. Hence, DFORL is easy to use and does not need a strong language bias. We demonstrate that DFORL can perform well on several standard ILP datasets, knowledge graphs, and probabilistic relation facts and outperform several well-known differentiable ILP models. Experimental results indicate that DFORL is a precise, robust, scalable, and computationally cheap differentiable ILP model.

We present an approach to automatically synthesise recursive predicates in Separation Logic (SL) from concrete data structure instances using Inductive Logic Programming (ILP) techniques. The main challenges to make such synthesis effective are (1) making it work without negative examples that are required in ILP but are difficult to construct for heap-based structures in an automated fashion, and (2) to be capable of summarising not just the shape of a heap (e.g., it is a linked list), but also the properties of the data it stores (e.g., it is a sorted linked list). We tackle these challenges with a new predicate learning algorithm. The key contributions of our work are (a) the formulation of ILP-based learning only using positive examples and (b) an algorithm that synthesises property-rich SL predicates from concrete memory graphs based on the positive-only learning. We show that our framework can efficiently and correctly synthesise SL predicates for structures that were beyond the reach of the state-of-the-art tools, including those featuring non-trivial payload constraints (e.g., binary search trees) and nested recursion (e.g., n -ary trees). We further extend the usability of our approach by a memory graph generator that produces positive heap examples from programs. Finally, we show how our approach facilitates deductive verification and synthesis of correct-by-construction code.

Many domains in the field of Inductive Logic Programming (ILP) involve highly unbalanced data. A common way to measure performance in these domains is to use precision and recall instead of simply using accuracy. The goal of our research is to find new approaches within ILP particularly suited for large, highly-skewed domains. We propose Gleaner, a randomized search method that collects good clauses from a broad spectrum of points along the recall dimension in recall-precision curves and employs an “at least L of these K clauses” thresholding method to combine sets of selected clauses. Our research focuses on Multi-Slot Information Extraction (IE), a task that typically involves many more negative examples than positive examples. We formulate this problem into a relational domain, using two large testbeds involving the extraction of important relations from the abstracts of biomedical journal articles. We compare Gleaner to ensembles of standard theories learned by Aleph, finding that Gleaner produces comparable testset results in a fraction of the training time.

Proposes a method for algorithmic learning of transmembrane domains based on elementary formal systems. An elementary formal system (EFS) is a kind of a logic program consisting of if-then rules. With this framework, the authors have implemented the algorithm for identifying transmembrane domains in amino acid sequences. Because of the limitations on computational resources, they restrict candidate hypotheses to EFSs defined by collections of regular patterns. From 70 transmembrane sequences and a similar amount of negative examples which are not transmembrane sequences, the algorithm has produced several reasonable hypotheses of small size. Experiments with the database PIR show that one of them recognizes 95% of 689 transmembrane sequences and 95% of 19256 negative examples which consist of non-transmembrane sequences of length around 30 randomly chosen from PIR.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Inductive Logic Programming (ILP) is an emerging research area at the intersection of machine learning, logic programming and software engineering. The first workshop on this topic was held in 1991 in Portugal (Muggleton, 1991). Subsequently, there was a workshop tied to the Future Generation Computer System Conference in Japan in 1992, and a third one in Bled, Slovenia, in April 1993 (Muggleton, 1993). Ideas related to ILP are also appearing in major AI and machine learning conferences and journals. Although European-based and mainly sponsored by ESPRIT, ILP aims at becoming equally represented elsewhere; for example, among researchers in America who are investigating relational learning and first order theory revision (see, for example, the papers in Birnbaum and Collins, 1991) and within the computational learning theory community. This year's IJCAI workshop on ILP is a first step in this direction, and includes recent work with a broader range of perspectives and techniques.

Though inductive logic programming (ILP for short) should mean the “induction of logic programs”, most research and applications of this area are only loosely related to logic programming. In fact, the automatic synthesis of “true” logic programs is a difficult task, since it cannot be done without a lot of information on the sought programs, and without the ability to describe in a simple way well-restricted searching spaces. In this chapter, we argue that, if such knowledge is available, inductive logic programming can be used as a valid tool for software engineering, and we propose an integrated framework for the development, maintenance, reuse, testing, and debugging of logic programs.

Inductive logic programming (ILP) realizes inductive machine learning in computational logic. However, the present ILP mostly handles classical clausal programs, especially Horn logic programs, and has limited applications to learning nonmonotonic logic programs . This article studies a method for realizing induction in nonmonotonic logic programs. We consider an extended logic program as a background theory, and introduce techniques for inducing new rules using answer sets of the program. The produced new rules explain positive/negative examples in the context of inductive logic programming. The proposed methods extend the present ILP techniques to a syntactically and semantically richer framework, and contribute to a theory of nonmonotonic ILP.