基于DNA粒计算的文本文献知识颗粒化方法研究

Research on a Text Document Knowledge Granulation Method Based on DNA Granular Computing

随着图书馆文献数量越来越多,其大数据化的趋势越来越明显,与此同时,读者“精准”获取所需文献内容或信息的难度加大,这是图书馆文献或信息检索服务面临的挑战。以超级基因、粒计算和DNA计算理论为基础,提出了一种文本文献知识颗粒化的新方法,其中包括:文本颗粒化定义,字符串寻址、抽取和组装定义,基因变异算子,摘要生成方法设计。实证包括3篇知网论文、3段书籍章节和3篇百度新闻的摘要生成,实验证明该方法可行。比较发现,有别于图书馆文献检索整体下载服务和知识元检索服务,该方法为图书馆提供了一种新的文献检索服务模式,其研究具有一定的学术参考价值和实际应用价值。

With the increase of library documents,the trend of big data is becoming more and more obvious.At the same time,it is more difficult for readers to“accurately”obtain the required literature content or information.This is a challenge for literature or information retrieval services of library.Based on the theories of super gene,granular computing and DNA computing,this paper proposes a new method of text document knowledge granulation,including a text granulation definition,a definition of string addressing,extraction and assembling,and the design of gene mutation operator and abstract generation method.The empirical research includes three CNKI papers,three book chapters and three Baidu news abstracts.The experiments show that this method is feasible.After comparison,it is found that this method provides a new document retrieval service mode,which is different from the overall document retrieval download services and knowledge element retrieval services of library,and this research has a certain academic reference value and practical application one.