Hierarchy Encoding Based XML Query Estimation

As the huge amount of XML data emerging in Web, researchers begin to focus on the topic of querying against XML data efficiently. Different from the query processing in relational data, XML query is characterized by structural join. Hence, quickly evaluating the cost of the join operation with the size of the intermediate results is an important part of XML query plan generation In this paper, an efficient hierarchy encoding scheme, H-encoding, is designed, and a model based on it for quickly estimating the size of the intermediate results is proposed This model provides accurate estimation for common containment (ancestor-descendent) query patterns, and at the same time it can deal with a class of order-axis query patterns that few research has covered. Some primary experiments on the H-encoding scheme are provided.