Research For Information Matching Technology Based On The Standards Of Audit Data

With the development of the information technology, traditional audit method cannotfit the requirement of audit business at present; In that case, it is an effective method thatuses computer to complete the audit work. When we use the computer to complete the auditwork, we should firstly deal with the audit data. Since the audited entity cannot provide thestandard audit data, we must match the audit data. Therefore, in this thesis, we study theinformation matching technology basing on the standard of the audit data, and use thistechnology to help the audit person to complete the data matching.The operating of the data matching is done by audit staff in the past. Although usingmanual method can complete the data matching accurately, it also cost much time. In thisthesis, we propose an audit information matching method, it can help audit person tocomplete audit work. The method is made up of two parts. One is tables matching method,the other is fields matching method. The tables matching method mainly uses first-orderlogic and semantic similarity technology to implementation the tables matching process.The steps of the method includes：Table matching method based on first-order logic, tablematching method based on semantic similarity and table matching merging results. Thefields matching method includes four steps. Firstly, we transform fields into the form ofstandardization; secondly, we use classification model to classify fields; thirdly, we usegeneration and feedback algorithm to match fields; finally, we use semantic similarity tofind the matching pair. The steps of the method includes：The processing of the socialsecurity data, the processing of the classification model, generation and feedback algorithmbased on the standard of audit data and matching selection based semantic similarity.Finally, we verify the validity of the algorithm by experiments, and we analysisexperimental results. The result verifies the correctness and efficiency of the method.