Research On XML-based Mobile Information Retrieval Model

Recently, information retrieval takes much convenience to people's lives. And XML has widely been used in information management, E-commerce, mobile communication, E-learning, exchange of e-documents and etc. There are 4 sections in a model that focused on the research on XML-based mobile information retrieval, they are: the set up of index, weighting, retrieval and the computing of similarity.Though big advancement has been obtained in the domain, there are many problems still left. Such as: how to reflect the demand of users, how to develop precision and recall, and how to mining most of the information in content and structure of XML documents.In order to solve these problems, studying and analyzing XML and vector space, information retrieval theories deeply are very important, and to blend context database, position of structure and distance of node into XML-based information retrieval model, then a method includes context database weighting, structure position and node distance has been put forward. The summary design, realize and the implementation of related arithmetic are given.The contents are as follows:1 The structure designing of archetype system is developed, and its implement path and main models function also have been introduced.2 The set up of index database, and 3 methods for keywords weighting, and the computing of similarity between query and document, and the main arithmetic on research and mobile client server are all implemented in this paper.3 An XML-based information retrieval system is realized based on Java platform, with Servlet and JDBC to implement the data transfer and accessing to database.4 The recall-precision curve is analyzed and got the improved result.The aim is to solve some applied problems and in the mean while accumulate some experience for applying system. There are good application and science value in the XML-based mobile information retrieval model.