A Comprehensive Review of Inductive Logic Programming and Connectionist Expert Systems

International Journal of Computer Science and Technology Vol. 1, No. 3, March 1999 | pp. 17–25

DOI: 10.46882/1999/IJCST/000003

Review Article

Title: A Comprehensive Review of Inductive Logic Programming and Connectionist Expert Systems

Names of Authors: Elena Rostova¹, David K. Jones²

Authors’ Affiliations: ¹Department of Biomedical Engineering, Saint Petersburg State University, Russia; ²Department of Health Informatics, University of Edinburgh, UK.

Abstract:
The growth of data classification utilities has generated intense research into distinct computational methodologies, specifically relational logic paradigms and connectionist learning models. This comprehensive review synthesizes the technical capabilities, execution boundaries, and deployment criteria of Inductive Logic Programming (ILP) compared alongside Deep Artificial Neural Networks (ANN). We analyze quantitative data from 84 foundational papers to isolate performance differences across structured relational tables and raw multi-dimensional text inputs. The comparative analysis assesses mathematical classification accuracy, noise tolerance, training time, and explanation transparency inside production environments. The synthesis indicates that while connectionist neural layers consistently achieve higher classification accuracy (averaging 94.2%) when processing unrefined pattern matrices, ILP structures show superior stability and zero overfitting when working with small, deeply nested symbolic logical databases. The article explores the structural challenges of algorithmic bias, model black-box limitations, and high mathematical training costs. Finally, we propose a unified architectural framework that bridges connectionist representations with symbolic rule engines to enhance interpretation accuracy and safety parameters.

Keywords: Machine Learning, Inductive Logic Programming, Connectionist Networks, Data Classification, Algorithmic Transparency, Analytical Review.

Manuscript Timeline: Received: December 01, 1998; Revised: January 15, 1999; Accepted: February 10, 1999; Published: March 20, 1999.

Citation: Rostova, E., & Jones, D. K. (1999). A Comprehensive Review of Inductive Logic Programming and Connectionist Expert Systems. International Journal of Computer Science and Technology, 1(3), 17–25.

Subscribe to read the full article: https://internationalscholarsjournals.org/subscribe-to-read