【摘要】 针对人工标引中存在的问题,提出了关键词自动标引应在增量、组合、排序三个方面对人工标引的结果加以优化。以"核反应堆工程"领域的期刊论文为语料开展实证研究,通过引入知识组织工具,利用字符串模式匹配法自动抽取候选词,按照比例归一化方法赋权值,设置一定的入口条件,以获取足量、优质、有序的标引词。实验结果表明,利用该方法进行关键词自动标引,有助于提高关键词标引的质量。更多还原

【Abstract】 To avoid the problems in keywords artificial indexing, keywords automatic indexing should be optimized in num-ber, combination form and sequencing. An empirical research is carried out based on the corpus of the papers on "NuclearReactor Engineering". It tries to get an appropriate number of high quality and orderly index terms by introducing knowl-edge organization tools, extracting automatically candidate keywords using string matching method, evaluating according tothe proportion normalization method and setting several entrance conditions. Experimental results show that this method iseffective to improve the quality of keyword indexing.更多还原