半结构化、层次数据的模式发现

Web数据资源及数据集成引发了半结构化数据问题，半结构化数据指其结构隐含或不规整的自描述数据。由于缺乏独立于数据的模式，有效地查询划浏览该类数据比较困难，半结构化数据的模式发现成为解决该问题的基础步骤。本文提出的算法能够快速有效地发现半结构化层次数据中的规整结构。它采用自顶向下的生成，结合有效的剪枝策略，从OEM模型表达的半结构化层次数据中构建模式树。

SCHEMA DISCOVERY FOR SEMISTRUCTURED HIERARCHICAL DATA

Semistructured data arise when the source does not impose a rigid structure (such as the web) and when data is combined from several heterogeneous data sources. Semistructured data is a self describing data whose structure is implicit or irregular. The lack of external schema information makes querying and browsing these data inefficient. This paper presents an algorithm discover schema fastly and efficiently. Using a top down approach with efficient prune strategy, a schema tree can be constructed from an OEM graph.