【摘要】 采用基于社交实值的反向传播神经网络和SimRank⁺⁺拓扑模型,对数据集学习分析得到完整的个性化推荐系统。为直观地看到分析结果,实验过程将搭建一个模拟社交网站的页面来呈现预测结果,通过在同一数据集上进行验证,并与现在主流的各大推荐算法进行对比,给出该研究算法的精确度和预测速度,从而验证利用社交实值条件推荐算法的有效性。更多还原

【Abstract】 By using the back propagation neural network based on social real value and SimRank⁺⁺ topology model, a complete personalized recommendation system for data set learning analysis is obtained. In order to visually see the results of the analysis, a simulation of social website is built in the experimental process to show the forecasting results. By comparing with the main recommendation algorithms on the same data set, the accuracy and prediction speed of the proposed algorithm are given to verify the effectiveness of the proposed algorithm.更多还原