The Release Of Military News Map And Research On Case Retrieval Method

With growingly develop and increasingly serious of international situation, there are a variety of news would be reported by the media every day. But because of the number of news is various and the isolation between the news media, readers often need to spend more time and energy to choose and retrieve.In view of this problem, through gathering and concluding the international military news of three mainstream media ("Sina", "Souhu" and "Phoenix"), and introducing "Baidu" map to locate and demonstrate the accident position of news, the paper has developed a unified platform to integrate different sources of news events. By integrating case-based reasoning technique, the platform regards the news as certain part of one case, such as its causes, development processes, results and influences, etc. and provides related function from news to cases. Exceptionally, it integrates case-based reasoning technique, proposes a case retrieval method based on international military, which can not only improve the efficiency of information retrieval, but also make the research results more accurate and perfect by the semantic analysis of keywords.This article analyzes the background and status of researching both at home and abroad on the news map and case-based reasoning firstly, expresses the richness of news map data and powerful data analysis capabilities of case-based reasoning. Aiming at the characteristics of the both, by some reasonable integration, the article develops a case retrieval method based on military news, which, on the one hand improves the application level of the media data, and makes news have the function of map location and event reasoning, besides the traditional function of information dissemination, on the other hand, also solves the problem of lacking of data source in case-based reasoning by integrating the military news data with a common theme into cases. Then it makes a detailed introduction respectively about the news data collection, case management, construction of news ontology tree analysis and research on case retrieval algorithm. Finally, combining the concept of recall, precision and F-test value, the paper does a performance experiment for the semantic matching based on ontology and traditional character matching, the result of which shows that the improved algorithm has greatly enhanced in retrieval performance because it fully considers the intrinsic relationship between words.