云计算环境下面向位置服务的多源信息融合方法研究——基于RBF神经网络仿真视角

Research on the Method of Multi-source Information Fusion for LBS in Cloud Computing Based on RBF Neural Network

云计算网络环境中,用户创作信息量急剧增长且凸显出明显的位置特征,针对如何充分发掘信息潜藏价值为用户决策提供服务支撑,本文应用RBF神经网络融合原理,提出面向位置服务的多源信息融合方法;应用空间向量模型,在对多用户创作位置信息进行特征维度提取基础上,对用户创作位置信息从不同维度进行业内访谈评分及网络训练,建立仿真网络对测试数据进行仿真实验,通过对比实际访谈数据,仿真误差低于0.35,研究表明该融合方法具有较高准确度。

In cloud computing,the amount of user generated content（UGC）is increasing sharply and the character of location is more and more obvious.For the decision support by mining the location character of information,this paper proposes a method of multi-source information fusion for location-based service（LBS） by applying a RBF neural network.Based on extracting the characteristic dimension for the location information created by multi-users,we apply Vector Space Model（VSM）,score the information from different dimensions through the interviews with the industry professionals and do the network study.We do the simulation experiment,and compare the test data with the one from actual interview.The simulation error is less than 0.35.The research shows that the fusion method has a higher accuracy.