Special Issue
Special Issue
The Role of Ontologies and Knowledge in Explainable AI
The Role of Ontologies and Knowledge in Explainable AI
Explainable AI (XAI) has been identified as a key factor for developing trustworthy AI systems. The reasons for equipping intelligent systems with explanation capabilities are not limited to user rights and acceptance. Explainability is also needed for designers and developers to enhance system robustness and enable diagnostics to prevent bias, unfairness, and discrimination, as well as to increase trust by all users in why and how decisions are made.
The interpretability of AI systems has been described long time ago since mid 1980s, but until recently it becomes an active research focus in computer science community due to the advances of big data and various regulations of data protection in developing AI systems, such as the GDPR. For example, according to the GDPR, citizens have the legal right to an explanation of decisions made by algorithms that may affect them (e.g., see Article 22). This policy highlights the pressing importance of transparency and interpretability in algorithm design.
XAI focuses on developing new approaches for explanations of black-box models by achieving good explainability without sacrificing system performance. One typical approach is the extraction of local and global post-hoc explanations. Other approaches are based on hybrid or neuro-symbolic systems, advocating a tight integration between symbolic and non-symbolic knowledge, e.g., by combining symbolic and statistical methods of reasoning.
The construction of hybrid systems is widely seen as one of the grand challenges facing AI today. However, there is no consensus regarding how to achieve this, with proposed techniques in the literature ranging from knowledge extraction and tensor logic to inductive logic programming and other approaches. Knowledge representation---in its many incarnations---is a key asset to enact hybrid systems, and it can pave the way towards the creation of transparent and human-understandable intelligent systems.
This special issue will feature contributions dedicated to the role played by knowledge bases, ontologies, and knowledge graphs in Explainable Artificial Intelligence (AI), in particular with regard to building trustworthy and explainable decision support systems. Knowledge representation plays a key role in Explainable AI (XAI). Linking explanations to structured knowledge, for instance in the form of ontologies, brings multiple advantages. It does not only enrich explanations (or the elements therein) with semantic information---thus facilitating evaluation and effective knowledge transmission to users---but it also creates a potential for supporting the customisation of the levels of specificity and generality of explanations to specific user profiles or audiences. However, linking explanations, structured knowledge, and sub-symbolic/statistical approaches raise a multitude of technical challenges from the reasoning perspective, both in terms of scalability and in terms of incorporating non-classical reasoning approaches, such as defeasibility, methods from argumentation, or counterfactuals, to name just a few.
The scope of this special issue is open to contributions by researchers, from both academy and industry, working in the multidisciplinary field of XAI.
Cognitive computational systems integrating machine learning and automated reasoning
Knowledge representation and reasoning in machine learning and deep learning
Knowledge extraction and distillation from neural and statistical learning models
Representation and refinement of symbolic knowledge by artificial neural networks
Explanation formats exploiting domain knowledge
Visual exploratory tools of semantic explanations
Knowledge representation for human-centric explanations
Usability and acceptance of knowledge-enhanced semantic explanations
Evaluation of transparency and interpretability of AI Systems
Applications of ontologies for explainability and trustworthiness in specific domains
Factual and counterfactual explanations
Causal thinking, reasoning and modeling
Cognitive science and XAI
Open source software for XAI
XAI applications in finance, medical and health sciences, etc.
Submissions shall be made through the Semantic Web Journal website. Prospective authors must take notice of the submission guidelines.
We welcome four main types of submissions: (i) full research papers, (ii) reports on tools and systems, (iii) application reports, and (iv) survey articles. While there is no upper limit, paper length must be justified by content.
Note that you need to request an account on the website for submitting a paper. Please indicate in the cover letter that it is for the "The Role of Ontologies and Knowledge in Explainable AI" special issue. All manuscripts will be reviewed based on the SWJ open and transparent review policy and will be made available online during the review process.
Also note that the Semantic Web Journal is open access.
It is worth noting note that submissions must comply with the journal’s Open Science Data requirements.
Paper submission: February 15, 2022 (Papers submitted earlier will be reviewed upon receipt)
Tentative date for Acceptance/rejection notification: June 30, 2022
Estimated Publication Date: October 2022
Free University of Bozen-Bolzano, Faculty of Computer Science, Italy
Free University of Bozen-Bolzano, Faculty of Computer Science, Italy
Department of Computing Science, Umeå University, Sweden
Free University of Bozen-Bolzano, Faculty of Computer Science, Italy
Research Centre in Intelligent Technologies (CiTIUS), University of Santiago de Compostela (USC), Spain
University of Plymouth, Faculty of Health, UK
The guest editors can be reached at ontologies-knowledge-in-xai-swj [at] googlegroups [dot] com