Incorporating domain knowledge in spatial clustering

将领域知识纳入空间聚类

• 批准号: 355996-2008
• 负责人: Wang, Xin
• 金额: $ 1.09万
• 依托单位: University of Calgary
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2012
• 资助国家: 加拿大
• 起止时间: 2012-01-01 至 2013-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=503189
• 关键词: Incorporating; domain; knowledge; spatial; clustering

Spatial clustering is an important topic in knowledge discovery research. It is to partition similar data objects into the same group (called cluster) based on their distance, density and reachability in space. It can be used to find natural clusters, to identify hot spots, and to partition an area based on utility. Ignoring domain knowledge and users' requirements during the spatial clustering can lead to useless and distorted clustering results. However, most existing clustering methods do not consider them during the clustering process, which prevents users from precisely describing their goals and understanding the results. Although some methods consider constraints, they only consider sharply limited knowledge concerning the domain and the users.

空间聚类是知识发现研究中的一个重要课题。它是将相似的数据对象根据它们在空间上的距离、密度和可达性划分为同一组（称为簇）。它可以用来寻找自然集群，识别热点，并根据效用划分区域。在空间聚类过程中忽略领域知识和用户需求会导致聚类结果的无用和失真。然而，大多数现有的聚类方法不考虑他们在聚类过程中，这妨碍了用户精确地描述他们的目标和理解的结果。虽然有些方法考虑了约束条件，但它们只考虑了关于域和用户的非常有限的知识。