Finding efficient and useful ways to search and index XML documents is a popular research topic in the field of computer and information science today. The path-based indexing method shows disadvantages of performance degradation when performing join operations of ancestor-descendent relationships and searching for middle and lower level nodes. To alleviate these disadvantages, a numbering scheme based indexing technique was proposed. This technique shows better performance in a variety of queries. However, a numbering scheme based indexing method is necessary to assign numbers to all nodes of all XML documents. It occurs the problem of both search overhead and disk space usage for indexes. In this paper, we propose a novel method that can efficiently construct and manage common paths of all XML documents. The proposed method stores similar structured XML documents more efficiently. In addition, it supports both insertion and deletion of XML documents more flexible.