Ability-oriented Performance Evaluation of General Intelligent Decision-making Engines Based on Knowledge Graphs

• Published in: 2021 IEEE International Conference on Software Engineering and Artificial Intelligence (SEAI) pp. 48 - 52
• Main Authors: Xu, Xin, Gao, Yan, Li, Xiyu
• Format: Conference Proceeding
• Language: English
• Published: IEEE 11.06.2021
• Subjects: ability-oriented; Cognition; Decision making; decision-making engine; evaluation; Hazards; Knowledge discovery; knowledge graph; military; Performance evaluation; Semantic search; Software algorithms
• DOI: 10.1109/SEAI52285.2021.9477540

The rapid development of artificial intelligence technology has changed modern warfare, in which intelligent decision-making based on knowledge graph technology is one of the most important applications. A general intelligent decision-making engine is expected to realize intelligent semantic search, intelligent question answering, and knowledge reasoning by building knowledge graphs from big data to improve the effectiveness of decision-making in military scenarios. However, there is still no answer on how to evaluate the performance of knowledge graph techniques serving for decision-making. For this reason, an ability-oriented performance evaluation framework for general decision-making engines used in the military field is proposed in this paper. We summarize five basic abilities of general intelligent decision-making engines from its construction and application with military knowledge graphs, and then discuss quantitative evaluation methods for each specific ability in detail.