近似图包含搜索和索引技术的研究

Study of similar graph containment search and index

在对图形数据库中的几种有代表性的传统相似性搜索及索引构造方法进行总结分析的基础上,探讨了近似图包含搜索区别于传统相似性搜索的特征,并且提出了一种针对近似图包含搜索的基于覆盖率和支持度对频繁子模式进行筛选的索引构造算法。实验结果验证了该方法的有效性。

Based on profound analysis and conclusion of several typical indexing strategies towards traditional search, the unique characteristics of similar graph containment search was discussed and a coverage and support based frequent pattern filtering approach for constructing index for this kind of search was proposed as well. Experimental results show this approach is effective.