讨论了有意义的非完全结构XML查询(NFS)结果的定义,提出了一种易于扩展的基于模式和实体的有意义判断模型——PE模型;基于PE模型,设计了具体的等价模式和等价查询项判断方法,提出了PE索引和I2P倒排索引,设计了一种有效的NFS查询算法来处理有意义的判断以及路径查询和关键字查询.实验表明,文中方法的查询质量和效率要优于XSEarch系统和Timber系统.
This paper discusses the issue of meaningful result determination for non-fully structured query (NFS) and proposes a scalable PE model to determine meaningful results based on the concept of pattern and entity. Within the framework of PE model, the paper proposes the method of identifying equivalent patterns and equivalent terms, puts forward the PE index and the improved inverted index I2P, and develops the effective and efficient NFS query algorithm for meaningful result evaluation, and the path query and keyword-based query. Experimental results indicate that the authors' approach outperforms XSEarch system and Timber system both on the querying quality and the efficiency of query processing significantly.