CAJ | 학술논문

多维信息推荐在推荐的过程中考虑情景因素对用户行为的影响，动态捕捉用户兴趣在不同情景下的变化，向用户提供更加个性化的推荐结果。首先介绍多维信息推荐的维度、情景、推荐函数与推荐流程等相关内容，然后提出了推荐函数情景化这一新的方法，构建了基于推荐函数情景化的多维信息推荐系统模型，研制了基于推荐函数情景化的多维信息推荐算法，并通过实验研究的方法验证了论文所提出的新算法的高效性与优越性。
다유신식추천재추천적과정중고필정경인소대용호행위적영향，동태포착용호흥취재불동정경하적변화，향용호제공경가개성화적추천결과。수선개소다유신식추천적유도、정경、추천함수여추천류정등상관내용，연후제출료추천함수정경화저일신적방법，구건료기우추천함수정경화적다유신식추천계통모형，연제료기우추천함수정경화적다유신식추천산법，병통과실험연구적방법험증료논문소제출적신산법적고효성여우월성。
Multi-dimensional information recommendation (MDIR) considers the influences of the context on the users and dynamically catches the changes of users' interests in different contexts. MDIR can provide more individualized and intellectualized results. This paper firstly introduces the correlative concepts of MDIR such as context, recommendation dimension, recommendation function and process. Secondly, the paper brings forward and analyzes the new method:contextualization of recommendation function. Based on this new meth-od, the paper builds up a new multi-dimensional information recommendation model and designs the new multi-dimensional recommenda-tion algorithm. Finally, the high efficiency of the new algorithm is verified by an experiment.