欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
随着满足用户需求的候选服务数量的飞速增长，服务选择的难度日益增大，服务推荐已成为服务选择的重要环节之一，受到越来越多的关注．然而，目前基于协同过滤的服务推荐方法较少关注到服务的不同属性特征对相似度计算会产生不同的影响，在寻找邻居用户时也很少考虑推荐用户与目标用户之间的信任关系，难以抵抗推荐用户的恶意推荐，无法保障推荐结果的精确度和町信性．针对以L问题，文中通过引入服务的推荐属性特征，改进传统相似度计算公式并基于Beta信任模型建立用户间信任关系，根据改进的相似度计算方法与服务推荐行为的信任度构建出邻居用户的可信联盟，提出了一种基于可信联盟的服务推荐方法．仿真实验与结果表明：与现有其它诸多方法相比，该方法不仅提高了服务推荐的精确度，还能有效保障服务推荐者的可信性，能较好的抵抗恶意攻击．
英文摘要：
With rapid amount growth of Web Services which can meet the requirements of users, service selection has become more and more difficult than ever. As a result, service recommenda tion has become one of the crucial aspects to service selection. More and more attention has been paid on service recommendation. However, most of the current collaborative filtering-based recommendation methods seldom recognize different impacts of different service attribute character- istics upon the calculation of similarity between users; neither do they pay much attention to the trust relationship between recommending and target users when finding neighbor users. There- fore, these methods cannot effectively counter against malicious attack from recommending users, and accuracy and trustworthiness of recommendation results cannot be guaranteed either. To address these problems, this paper modified the traditional similarity calculation method with an introduction of service recommendation attribute characteristics, and employed Beta trust model for the establishment of trust relationship between users. Based on the combination of improved similarity calculation method and the trust degree of service recommendation behavior, trustworthy community was set up to find neighbors, followed with a service recommendation method based on trustworthy community （SRMTC）. Simulation and experiment results demonstrated that compared with most other service recommendation methods, SRMTC not only increased the accuracy of recommendation results but also guaranteed the trustworthiness of service reeom menders and could better counter against malicious attack.
同期刊论文项目
同项目期刊论文
Improved Virtual Potential Field Algorithm Based on Probability Model in Three-Dimensional Direction
Image Hiding Algorithm in Discrete Cosine Transform Domain Based on Grey Prediction and Grey Relatio
Multi-sample clustering decision fusion for distributed target detection in wireless sensor networks
期刊信息
位置：