Despite recent advances in modern machine learning algorithms, the opaqueness of their underlying mechanisms continues to be an obstacle in adoption. To instill confidence and trust in artificial intelligence (AI) systems, explainable AI (XAI) has emerged as a response to improve modern machine learning algorithms’ explainability. Inductive logic programming (ILP), a subfield of symbolic AI, plays a promising role in generating interpretable explanations because of its intuitive logic-driven framework. ILP effectively leverages abductive reasoning to generate explainable first-order clausal theories from examples and background knowledge. However, several challenges in developing methods inspired by ILP need to be addressed for their successful application in practice. For example, the existing ILP systems often have a vast solution space, and the induced solutions are very sensitive to noises and disturbances. This survey paper summarizes the recent advances in ILP and a discussion of statistical relational learning (SRL) and neural-symbolic algorithms, which offer synergistic views to ILP. Following a critical review of the recent advances, we delineate observed challenges and highlight potential avenues of further ILP-motivated research toward developing self-explanatory AI systems.
The increasing reliance of modern science in computer simulation demands appropriate trust in simulation models for credible results. Because of its foundations in operations research, model credibility is conventionally viewed from the lens of numerical and transformational accuracy. However, the exploratory use of models in scientific discovery, causal explanation, and strategic decision-making render this view incomplete. Recognising the significance of the cognitive interests of model users and the context-sensitive, adaptive nature of building confidence in scientific models, we characterise credibility as trust. Appropriate and justifiable trust is conceptualised as a dynamic, cognitive construct that evolves through interactive, experiential learning. Following the delineation of the dimensions and attributes of trust, conceptual foundations of a dynamic trust model, including alternative measurement strategies, are proposed. Guidelines for trustable models are elaborated to provide a basis for exploiting synergies between the cognitive models of trust and model evaluation.
15 June 2021 in IEEE Transactions on Emerging Topics in Computational Intelligence
Human-robot interactive decision-making is increasingly becoming ubiquitous, and trust is an influential factor in determining the reliance on autonomy. However, it is not reasonable to trust systems that are beyond our comprehension, and typical machine learning and data-driven decision-making are black-box paradigms that impede interpretability. Therefore, it is critical to establish computational trustworthy decision-making mechanisms enhanced by interpretability-aware strategies. To this end, we propose a Trustworthy Decision-Making (TDM) framework, which integrates symbolic planning into sequential decision-making. The framework learns interpretable subtasks that result in a complex, higher-level composite task that can be formally evaluated using the proposed trust metric. TDM enables the subtask-level interpretability by design and converges to an optimal symbolic plan from the learned subtasks. Moreover, a TDM-based algorithm is introduced to demonstrate the unification of symbolic planning with other sequential-decision making algorithms, reaping the benefits of both. Experimental results validate the effectiveness of trust-score-based planning while improving the interpretability of subtasks.
31 January 2012 in Expert Systems with Applications