摘要
为提高关键词提取的准确度,提出一种基于用户反馈的关键词提取方法,利用用户反馈来修正关键词提取中产生的误差。将用户对抽取结果的反馈信息记录在用户反馈库中,通过用户反馈值来刻画用户对关键词的评价;在多次用户反馈后,建立用户对关键词提取的模型;结合统计特征信息和用户反馈模型来计算候选词的权重,基于权重值大小提取文本关键词。实验结果表明,该方法能在用户反馈中不断学习,提取的关键词具有较高的准确率。
A keywords extraction method based on user feedback was proposed to improve the accuracy of the keyword extrac- tion. The user feedback was used to correct the error generated in the keyword extraction. User feedback library recorded user feedback on the results of extraction, user feedback value was used to describe the evaluation of the user key words. After get- ting multiple user feedbacks, the keyword extraction model was established. The weight of the candidate words was determined by linearly combining the statistical characteristics of information and user feedback value, keyword extraction was generated ac- cording to the weight. The experimental results show that this method can study in the user feedback and it is better in evalua- tion index.
出处
《计算机工程与设计》
北大核心
2015年第8期2277-2281,共5页
Computer Engineering and Design
基金
国家自然科学基金项目(61272109)
关键词
关键词提取
用户反馈
用户反馈库
用户反馈值
统计特征
keyword extraction
user feedback
user feedback library
user feedback value
statistical characteristics
