Leveraging Explainable AI with Semantic AI: European Big Data Value Forum
Semantic AI and Explainable AI have been burgeoning fields within the software industry. In a recent discussion at the European Big Data Value Forum held in Vienna, a panel of experts in the field; including Semantic Web Company CEO Andreas Blumauer, sat down to discuss the importance of data-driven innovation within a data-centric economy. The panel, moderated by Semantic Web Company CFO Martin Kaltenböck, expressed concerns about whether self-learning algorithms were considered as “black boxes” for most non-AI experts. The discussion led to the conclusion that current AI solutions:
- Do not provide useful results
- Provide applications that are not fulfilling the requirements
- Make it very difficult to explain the processes that have led to a certain outcome or decision
As a result, the panel discussion elaborated on the benefits of using Semantic AI, which seeks to provide an infrastructure that overcomes information asymmetries between the developers of AI systems and other stakeholders, including consumers and policymakers. Semantic AI also helps to leverage enterprise AI governance, that is based upon three layers: technically, ethically, and on the social and legal layer.
According to the article published by Gasser and Almeida in 2017, the technical layer refers to the algorithms and data out of which AI systems are built upon. In addition to the technical layer, ethical concerns apply to all types of AI applications and are based on human rights principles. Finally, the social and legal layers address the process of creating institutions and allocating responsibilities for regulating AI.
During the panel discussion, Andreas Blumauer expressed the following:
“Large corporations such as Oracle and Google are using Enterprise Knowledge Graphs to have a well-developed semantic data infrastructure which is the basis for machine learning and explainable AI, knowing where your data comes from”.
