Symbolic Knowledge Extraction and Injection with Sub-symbolic Predictors: A Systematic Literature Review

2024-02-08

ACM Computing Surveys

AI Summary

The “black-box” nature of machine learning models severely limits their interpretability and trustworthiness, particularly in high-stakes domains. Method: This paper conducts a systematic study of Symbolic Knowledge Extraction (SKE) and Symbolic Knowledge Injection (SKI), grounded in a Systematic Literature Review (SLR) and an eXplainable AI (XAI) analytical framework. Contribution/Results: We propose the first unified meta-model and a dual-dimensional taxonomy—categorizing 132 SKE and 117 SKI methods along dimensions of objective, operation, input/output, and adapter type—thereby bridging a critical gap in cross-research between SKE and SKI. We further introduce a novel collaborative classification framework that generalizes and integrates existing surveys. All methods are annotated with their open-source implementation status to facilitate practical adoption and identify research gaps. Our work provides both theoretical foundations and practical tools for enhancing transparency and enabling trustworthy deployment of sub-symbolic predictors.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Cognitive Modeling & Cognitive Systems: Symbolic Representations

Application Category

• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In this article, we focus on the opacity issue of sub-symbolic machine learning predictors by promoting two complementary activities—symbolic knowledge extraction (SKE) and symbolic knowledge injection (SKI)—from and into sub-symbolic predictors. We consider as symbolic any language being intelligible and interpretable for both humans and computers. Accordingly, we propose general meta-models for both SKE and SKI, along with two taxonomies for the classification of SKE and SKI methods. By adopting an explainable artificial intelligence (XAI) perspective, we highlight how such methods can be exploited to mitigate the aforementioned opacity issue. Our taxonomies are attained by surveying and classifying existing methods from the literature, following a systematic approach, and by generalising the results of previous surveys targeting specific sub-topics of either SKE or SKI alone. More precisely, we analyse 132 methods for SKE and 117 methods for SKI, and we categorise them according to their purpose, operation, expected input/output data and predictor types. For each method, we also indicate the presence/lack of runnable software implementations. Our work may be of interest for data scientists aiming at selecting the most adequate SKE/SKI method for their needs, and may also work as suggestions for researchers interested in filling the gaps of the current state-of-the-art as well as for developers willing to implement SKE/SKI-based technologies.

Problem

• Machine Learning
• Model Interpretability
• Predictive Models

Innovation

• Symbolic Knowledge Extraction
• Symbolic Knowledge Injection
• Machine Learning Interpretability