Design And Implementation Of Archive Management System For Shenyang SIASUN Company

Back in the 1930s, our country systematic conducted research on file retrieval discipline. Closely integrated with the rapid development of information technology and the archival work, creating a steady stream of powerful driving force for research subjects to retrieve files. Nowadays, rapidly evolving information technology makes network-based file retrieval means for the deepening of, Traditional manual information retrieval will be gradually replaced by computer information retrieval.The key word is the most commonly used method of computer archival information retrieval method, which is applied to most people, but the retrieved information is largely broader and unfocused. Therefore, the system uses this method and natural language processing technology combined, through calculation method of text similarity between retrieving information and retrieved information, strengthen targeted retrieval, allowing the system to achieve better search results. This system applied the following techniques:1. The system uses natural language processing technology segmentation, the catalog and user submitted content keyword query string and text files of unknown origin were pretreated word.2. The system in the search process, the need for compute-intensive pretreatment words, to improve the calculation speed retrieval, using the inverted index storage structure in the data storage.3. In a variety of text similarity calculation method, the system uses the information retrieval system commonly used algorithms BM25. By a large number of experiments show good performance BM25 calculation method, and the relatively stable.After a lot of experiments in this system as well as a comprehensive and detailed test, the system has been working in the Shenyang SIASUN company, the system running fast and stable, the effect is significant archive retrieval.