Research On Ontology-based Approximate Query In XML Documents

XML (Extensible Markup Language) has been increasingly used in Web applications and becomes the standard of data interchange over the Internet. During the query processing of XML data, users’intents are often ambiguous and incomplete, so that users cannot express their purposes accurately. Thus results’ precision and recall rates are not satisfactory, the situation may be extreme when adding those data that do not meet the intent of the user’s query to the query result collection, resulting in too many results and losing the effective information; the requirements should be included in the retrieved data are ignored, resulting in the returned information is too small; misunderstanding the focus of user queries, which makes the results have a greater deviation with the initial intents of the users. A "null result" of the problem is divided into the following areas:the path does not match the organizational structure of the XML document; the naming rules between the given query and the data in XML documents are different; the query conditions are too strict, resulting in few answers meeting the query conditions; because there is no common understanding of domain knowledge for the users, the results meeting the query cannot be returned to the users.There are many methods to solve the problem of returning the null results, and the introduction of domain ontology is one of them. In order to solve the null result problem, domain ontology which is used to represent the semantics, and ontology mapping clustering methods are introduced to expand the query. To achieve a clear semantics, we often need to support of two types of semantic data:describing a specialized domain knowledge and providing a shared vocabulary to support the body; containing explicit semantic information in the document instance, that is, the body of ontology instances. Hidden from the XML document to extract semantic information to construct an XML document describes the formal semantic description of the body, which can be described in XML syntax layer information from the upgrade to the semantic layer.To meet the above requirements, this paper presents an approximate algorithms based on the ontology. The method is divided into three parts:Firstly, the parsing of XML, including the document elements, attributes and values, etc. isolated, extracted concepts and relationships between concepts, mapped to ontology concepts, attributes and relationships, to build a standard domain ontology, the overall the embodiment of semantic information in XML documents. Secondly, different methods of conflict to build multiple heterogeneous ontology, calculated by the similarity-based ontology mapping, semantic expansion of query terms. Then the structure of XML queries and conditions are extended, during the processing, firstly conditions are separated into multiple elements, in order of the importance for the relaxation of the selected elements to the ontology-based semantic similarity of query relaxation, the final selection based on relevance to the relaxation of the results.