A Pattern-based Approach For Extracting Linked Data From Chinese Government Documents

Recently, Linked Open Government Data (LOGD) is becoming an important way topromote better community participation, business development, and governmental trans-parency. However, there are few methods proposed to generate LOGD based on Chinesegovernment documents. To address this issue, we propose an Pattern-based approach thatautomatically extracts linked data from Chinese government document that consists ofgovernment rules. We proposes an innovative algorithm which generates patterns withoutrelying on a predefned knowledge base such as WordNet. The experimental results showthat the proposed method has a relative high performance in terms of precision and recall.In addition, we develop a search algorithm based on the generated linked data. The ex-perimental results indicate that our search algorithm has a much better performance thanthe keyword search method.Our contribution is as follows:Propose an innovative algorithm which extracts linked data from Chinese govern-ment document.Automatically extract linked data from Chinese government document.Our approach extracts linked data without relying on a predefned knowledge basesuch as WordNet.Automatically generate pattern, according to the type of documents and the rela-tionship that user defnes.