An Online Integration Platform Of Oa Journals And Research On Retrieval Service Schema Automatic Extraction

At present, DeepWeb is a hot research field in the database area. Open Access journals as a DeepWeb resource is developing very fast by its excellence concept in acdemic sharing.However, OA journals over the internet and its"island"status worsening day after day. Traditional search engine is hard to index its hidden data. A way to solve it is to integrate the OA journal's retrieval service to estiblish an vitual digital resource space.The first problem is lack of a flexible, scalable, open access resources online search service integrated platform architecture model.Second is lack of a way to extract the OA journal retrieval service form the retrieval interface.The paper has do many researchs in those aspects.First of all, according to the features of the OA journal and the functional requirements of the unified retrieve service platform designed a flexible, scalable, open access resources online search service integrated platform architecture model.And defines each component and its work related functions and data exchange interfaces.Secondly, we propose an approach to extract the retrieval service schema of OA journals based on the idea of classification according to analyzing a number of OA journals'retrieval interfaces. We first make a deep analysis on the retrieval forms of OA journals, classify the forms by their features, and use the html document analyzing technology to analyze the classified attribute units. Then we construct the description model of retrieval service schema according to the semantic information of attribute units. On the basis of that, we design a storage structure of retrieval service schema based on XML.Finaly, based on those works, by the experiments on the Prototype system we developed, retrieval service model for the automatic extraction is analyzed and evaluated and analysis of the performance of the platform architecture model from the recall, response time, etc.