Research And Implementation On Corresponding Method Between Chinese Name Of Transportation Data And Standard Terminology

With the wide application of computer technology in the transportation industry, thetransportation industry is developing rapidly. To push forward traffic informatizationconstruction is not only the inevitable choice of modernization of traffic, but also animportant means of accelerating the development of transport, In the process ofinformatization construction of transportation industry, more and more need for a uniformdata standards to improve the standardization of traffic data information, This paper, underthe support of the program of "Transportation information data standards compliancetesting", researched the corresponding method of Chinese name of traffic information data,which is an important function module to develop traffic information data standardscompliance testing system, but also the basis for the realization of standardization andinformatization of traffic industry, and it has an important significance to advance trafficinformation construction.In this thesis, first of all, on the basis of the analysis of Chinese word segmentationtechnology, used statistical method based on the ICTCLAS system for segmentation,expanded of original word library of ICTCLAS, added into the professional words ofdescribing the field of transportation. Experiments show that, after treatment, the results ofthe system can meet the needs of this research, can carry on the reasonable segmentation forthe Chinese name of the traffic data. Then, this thesis extracted the feature word, andcalculated weights of feature vectors according to the method of TF-IDF. In this thesis,based on Chinese text preprocessing, researched several similarity methods, and design andimplementation of each algorithm was carried out respectively. Including edit distance, theimprovement of the traditional edit distance algorithm, based on contextual similarityalgorithm, as well as the comprehensive similarity calculation method which consideringthe common effect of combining the edit distance with context. Through analysis andcomparison of the experimental results of several algorithms, selected the comprehensivesimilarity calculation method as the final method to achieve the compliance testing ofChinese name of transportation data and standard terminology, this method can achieve the match and search of user data more accurately in the system.In this thesis, comprehensive similarity calculation method was applied on"compliance test" module of the "Transportation information data standards compliancetesting", the system can run more stably, the main function can be realized, had achievedrequirements when the project was designed.