基于字串切分统计词典的繁体中文拼写检错方法

针对繁体中文拼写检错的问题进行了研究,提出一种基于字串切分统计词典的检错方法。利用语料库中字串出现的频率信息作为检错依据,根据字串及其频率信息来建立统计词典,并设计了基于统计规则评判的检错算法。以SIGHAN7会议中文拼写校验任务中用于检错评测的1 000句测试集作为实验测试集,并与此会议提交的结果进行比较,实验结果表明,与基于复杂语言模型的检错方法相比,该方法在实现简单的同时也有很好的检错效果,获得了较高的准确率和精确率以及较低的误报率。

浅谈互联网信息挖掘技术

本文就互联网信息挖掘技术和对网络信息挖掘中的关键技术、系统流程进行了阐述。

Knowledge Automatic Indexing Based on Concept Lexicon and Segm-entation Algorithm

This paper is based on two existing theories about automatic indexing of thematic knowledge concept. The prohibit-word table with position information has been designed. The improved Maximum Matching-Minimum Backtracking method has been researched. Moreover it has been studied on improved indexing algorithm and application technology based on rules and thematic concept word table.