| 基于改进KNN的案例匹配模块的设计与实现 |
| |
| 引用本文： | 谢开池,薛醒思.基于改进KNN的案例匹配模块的设计与实现[J].福建工程学院学报,2017,0(4):349-357. |
| |
| 作者姓名： | 谢开池 薛醒思 |
| |
| 作者单位： | 福建工程学院信息科学与工程学院 |
| |
| 摘 要： | 为了提高KNN检索策略的检索效率和检索结果的质量，提出一种改进的KNN检索策略。在引入图书馆领域本体和概念语义相似度度量技术的前提下，利用句法结构筛选不合理的案例以降低计算规模,从而提高案例的检索质量和效率，利用改进的微粒群算法优化概念语义相似度度量技术中的组合参数以提高KNN检索的结果质量。实验数据采用福州晓锋科技信息咨询有限公司提供的图书馆参考咨询测试数据。实验结果表明，相比于传统KNN和基于传统PSO的改进KNN方案有效地提高了案例匹配结果的查全率和查准率。
|
| 关 键 词： | 案例推理 KNN 微粒群算法 |
Design and implementation of a case matching module based on improved KNN |
| |
| Xie Kaichi,Xue Xingsi.Design and implementation of a case matching module based on improved KNN[J].Journal of Fujian University of Technology,2017,0(4):349-357. |
| |
| Authors: | Xie Kaichi Xue Xingsi |
| |
| Institution: | College of Information Science and Engineering, Fujian University of Technology |
| |
| Abstract: | To improve the efficiency and quality of case retrieval, an improved KNN retrieval strategy was proposed. By introducing library domain ontology and concept semantic similarity measurement technology, cases’ syntactic structure was employed to filter out the unreasonable cases to reduce the computation amount (search space) and improve the case retrieval (alignment’s) quality. Then, an improved particle swarm algorithm was presented to determine the optimal aggregating parameters in the similarity measure technologies to improve the case alignment’s quality. In the experiment, the testing cases were from Fuzhou Xiaofeng Science and Technology Information Consulting Ltd., Co,. The experimental results show that compared with the traditional KNN and the traditional PSO-based KNN, the proposal can significantly improve the case alignment’s quality in terms of both recall and precision. |
| |
| Keywords: | case based reasoning K nearest neighbourhood particle swarm algorithm |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《福建工程学院学报》浏览原始摘要信息 |
| 点击此处可从《福建工程学院学报》下载免费的PDF全文 |
