将面向属性的归纳方法(attribute—oriented induction)用于壁画的展示,提出一种基于知识发现的壁画展示方法。对壁画按内容、位置、时间等强相关维属性,引入本体的层次化描述方式用于对比展示,可帮助研究者更好地获取对象的隐性知识,启发新的类描述和关联规则的发现。结合基于绘画构图学特征的相关度评价方法,可有效地选取研究者关注的内容进行比较和展示。实验以真实的敦煌壁画研究课题为例,验证了本文方法在辅助壁画研究中的有效性。
By introducing attribute-oriented induction methodology (AOI) of data mining, this paper puts forward a new exhibition method for murals based on knowledge discovery. Retrieval results are organized according to the correlation dimension of murals for presentation of comparison, which is helpful for the researchers to acquire implicit knowledge and enlighten new association rules. We present a composition-based relevance ranking to display result images. The applications in Dunhuang mural demonstrate the effectiveness of our approach.