CAJ | 학술논문

针对传统的基于节点相似性的链接预测方法存在链接预测指标仅考虑网络结构信息或者节点属性信息，以及链接预测指标静态处理节点之间关系的问题，提出了一种基于信息融合相似性算法的链接预测指标（similarity based on network evolution and user generated content ， SNEUGC），该指标结合用户生成内容信息和网络演化信息对含权网络进行链接预测，以解决现有链接预测指标在含权网络环境下链接预测准确率低的问题。实验证明，该方法的准确率达到了80％，具有一定的可行性。
침대전통적기우절점상사성적련접예측방법존재련접예측지표부고필망락결구신식혹자절점속성신식，이급련접예측지표정태처리절점지간관계적문제，제출료일충기우신식융합상사성산법적련접예측지표（similarity based on network evolution and user generated content ， SNEUGC），해지표결합용호생성내용신식화망락연화신식대함권망락진행련접예측，이해결현유련접예측지표재함권망락배경하련접예측준학솔저적문제。실험증명，해방법적준학솔체도료80％，구유일정적가행성。
The traditional link prediction methods based on the similarity of nodes have several problems .Only the structure of network or the information of nodes were considered as link prediction indexes .In the prediction indexes the relationship be-tween two nodes was treated statically .A new link prediction index called SNEUGC on the basis of information fusion similarity algorithm was then proposed .The SNEUGC index makes link predictions in weighted networks considering both user generated content information and network evolution information , and can solve the problem of low prediction accuracy of those existed inde-xes in weighted networks .The result verifies that the accuracy of SNEUGC is up to 80%and the method is feasible .