摘要
本文首次将Fuzzy关系模式映射反演原则(FRPMI)应用于犯罪侦查自动推理中,提出了Fuzzy定映自动推理框架。通过实际犯罪模式与现场表现模式的关系结构,建立了两种模式映射。研究表明,在计算机中建立案件现场与破案线索的规范集是自动推理的前提,而补充线索是关键。作者认为,自动推理的可靠性决定于人机交互作用,同时破案线索的选择要通过综合评价来决定,并且不能缺少自学习和直觉与理性的判断。
Fuzzy relationship pattern mapping inversion (FRPMI) is firstly introduced to the criminal detection by automatic reasoning, and a frame of fuzzy mapping automatic reasoning is proposed. Through the relationship construction of actual crime model and on-the-spot model, a couple of mapping models are set up. The study shows that the premise of automatic reasoning is the sets of patterns of case on-the-spot and case-cracking clue, and the complementary of the clues is the key point. The reliability of the automatic reasoning depends on the man-computer interaction results. Simultaneously, selection of the case-cracking clue should be determined by comprehensive evaluations, and self-teaching and intuitional logical judgments are essentially needed.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2003年第1期70-75,共6页
Pattern Recognition and Artificial Intelligence
