CAJ | 학술논문

提出了一种基于关联规则的社交网络好友推荐算法，在进行好友推荐时，考虑现实社交活动中“志趣相投”的好友常常会关注相同的人和事，网络社交中的好友也常常会关注相同的“人”和“事”，将“关注”看成一条交易记录，把关注的用户看成交易项，所有交易项的集合看成交易数据库，生成二阶候选项集，并按支持数降序排序，推荐前犖个用户作为好友。以新浪微博993950条用户关注数据及552600条微博关注数据作为实验的对象，实验结果表明，算法具有良好的性能，可实现较高的召回率与准确率。
제출료일충기우관련규칙적사교망락호우추천산법，재진행호우추천시，고필현실사교활동중“지취상투”적호우상상회관주상동적인화사，망락사교중적호우야상상회관주상동적“인”화“사”，장“관주”간성일조교역기록，파관주적용호간성교역항，소유교역항적집합간성교역수거고，생성이계후선항집，병안지지수강서배서，추천전락개용호작위호우。이신랑미박993950조용호관주수거급552600조미박관주수거작위실험적대상，실험결과표명，산법구유량호적성능，가실현교고적소회솔여준학솔。
A social network friends recommendation algorithm based on association rules is proposed.Considering that in our daily life,the congenial friends always pay close attention to the persons and things of similar purpose and interests,so do cyber good friends,we take"focus"as a trade record,the focusing users as the trade terms and the set of all the trade terms as the trade da-ta,then the set of second order candidates is produced and we descend the number of supporters.The former N users are recom-mended as good friends.Taking 993 950 users’focus data and 552 600 focus data in Sina microblog as the experiment objects,we reach the conclusion that the proposed algorithm has good performance and can achieve the higher recalling rate and precision rate.