Reporter：Dr. Jun Liu, Ulster University, UK
Title of report 1：Knowledge-Centralized Decision Analytics under Uncertainty for Explainable AI
Report time：December 9, 2020 (Wednesday) 15:00 to 16:30 PM
Report location：Tencent Conference (Conference ID: 868 306 628) + Offline (Information Building 01020, Jiuli Campus )
Summary of report 1：Belief-rule based methodology (called BRB), as a novel advanced and generic rule-based decision model, is receiving increasing attention and is becoming increasingly affluent since its first publication. BRB fits well within Explainable AI (XAI) now that it is a white box approach which can provide direct access and transparency to decision makers and stakeholders. BRB methodology not only holds strengthens of the normal off-the-shelf rule-bases systems and even enhanced them further; it also resolved the limitations of the existing rule-based systems greatly. In addition, BRB decision model can be knowledge driven or data driven, the flexibility of which lends itself to be deployed in a wide range of applications in the last few years. This talk will highlight the key strengths of the BRB, especially clarify how the BRB along with its advancement can help resolve the limitations of the existing rule-based systems, then cover some of latest developments as well as identify the remaining challenges and some perspectives.
Title of report 2：Role of Automated Reasoning in Safe and Trusted AI
Report time：December 16, 2020 (Wednesday) 15:00 to 16:30 PM
Report location：Tencent Conference (Conference ID: 967 291 366) + Offline (Information Building 01020, Jiuli Campus )
Summary of report 2：Breakthroughs in artificial intelligence (AI) combined with data analytics and high-performance computing has seen the emergence of society-changing AI applications. There are some myths that AI is machine learning or deep learning, and symbolic AI techniques seem to be out of date. This talk focus on symbolic AI techniques, the foundation of AI and the key area in the AI, especially about automated reasoning (AR), highlighting its nature and the-state-of-the-art methods, followed by several high level application area illustrations in which AR can and have played very important roles. Accordingly, the role of AR for the cutting-edge research in ensuring the safety and trustworthiness of AI systems is briefly summarized.