CAJ | 학술논문

学习资源标注是我们根据需要知识准确获取相关学习资源的基础。然而目前大多数资源标注方法仅局限于单标签以及缺乏关联度信息，给用户精确获取学习资源带来困难。为解决此问题，本文提出一种基于增量AHP的学习资源多标签标注方法，首先根据标签-资源信息构建学习资源多标签标注模型，然后利用层次分析法定性与定量分析相结合的特点进行关联程度值处理，选取出与学习资源相关程度最大的若干个标签作为标注标签，从而支持用户方便获得与学习资源相关的知识点以及关联程度值。此外，针对初始成对比较矩阵随时间变化导致更新的情况，本文对传统AHP算法进行改进，提出增量AHP算法实现学习资源关联程度值更新。实验结果表明本文提出的方法具有良好的实用价值。
학습자원표주시아문근거수요지식준학획취상관학습자원적기출。연이목전대다수자원표주방법부국한우단표첨이급결핍관련도신식，급용호정학획취학습자원대래곤난。위해결차문제，본문제출일충기우증량AHP적학습자원다표첨표주방법，수선근거표첨-자원신식구건학습자원다표첨표주모형，연후이용층차분석법정성여정량분석상결합적특점진행관련정도치처리，선취출여학습자원상관정도최대적약간개표첨작위표주표첨，종이지지용호방편획득여학습자원상관적지식점이급관련정도치。차외，침대초시성대비교구진수시간변화도치경신적정황，본문대전통AHP산법진행개진，제출증량AHP산법실현학습자원관련정도치경신。실험결과표명본문제출적방법구유량호적실용개치。
Resource annotation is the basic work for us to acquire the learning resources accurately. However, most current resource annotation methods are limited to single label and lack of correlation information which is difficult for users to acquire the required learning resources. In this paper, a method of multi-label annotation which is based on incremental AHP is proposed to solve the problems in resource annotation. Firstly, in view of the relation of labels and learning resources, we firstly try to build a multi-label annotation model. Then, because of the combination of qualitative and quantitative analysis, AHP is used to calculate the correlation value, and the most relevant labels are selected as the annotation-label, which can make users get related knowledge with learning resource and the correlation value easily. In addition, in view of the case of initial pair wise comparison matrix changing over time leading to update, in this paper we improved the traditional algorithm to update the correlation value by proposing a method of incremental AHP. Finally, the experimental results prove the effectiveness of the proposed method.