On the integration of symbolic and sub-symbolic techniques for XAI: A survey

Issue title: Special Issue for the Twentieth Edition of the Workshop ‘From Objects to Agents’

Guest editors: Matteo Baldoni, Federico Bergenti, Stefania Monica and Giuseppe Vizzari

Article type: Research Article

Authors: Calegari, Roberta^{; *} | Ciatto, Giovanni | Omicini, Andrea

Affiliations: Dipartimento di Informatica – Scienza e Ingegneria (DISI), Alma Mater Studiorum–Università di Bologna, Italy

Correspondence: [*] Corresponding author: Roberta Calegari, Dipartimento di Informatica – Scienza e Ingegneria (DISI), Alma Mater Studiorum–Università di Bologna, Italy. E-mail: [email protected].

Abstract:  The more intelligent systems based on sub-symbolic techniques pervade our everyday lives, the less human can understand them. This is why symbolic approaches are getting more and more attention in the general effort to make AI interpretable, explainable, and trustable. Understanding the current state of the art of AI techniques integrating symbolic and sub-symbolic approaches is then of paramount importance, nowadays—in particular in the XAI perspective. This is why this paper provides an overview of the main symbolic/sub-symbolic integration techniques, focussing in particular on those targeting explainable AI systems.

Keywords: XAI, symbolic and sub-symbolic AI, explainability, interpretability, trustable system

DOI: 10.3233/IA-190036

Journal: Intelligenza Artificiale, vol. 14, no. 1, pp. 7-32, 2020