【摘要】 自然语言处理任务复杂,并且涉及庞大的数据量,需要在同一数据集上进行可重复的测试来比较不同研究方法的优劣。因此,技术评测一直都是自然语言处理技术发展的重要驱动。本研究从多维度调研了国内影响力较大的自然语言处理技术评测,并同国际上的评测语言资源在语料来源、语料规模、语料库建设方法等进行了对比分析,总结了国内评测语言资源存在的不足,分析了评测语言资源的建设方向和发展趋势。更多还原

【Abstract】 Processing natural language,a complicated task,which involves in enormous data amount,requires to repeat the test on the same data set to compare the adcvantages and disadvantages of the research methods. The test of technology,therefore,is aways the important drive for the development of the processing techonology of natural language. The research is to investigate the domestic influential test of technology from diverse dimensions and compare them with the international ones from the aspects of corpus sources,corpus size and the methods for the construction of the corpus so that the shortcomings of the domestic technology can be found to know the tendency of the technology.更多还原