众筹项目的社交网络影响力预测与分析

发布时间：2018-04-28 11:29

本文选题：社交网络 + 众筹项目　；参考：《西安交通大学学报》2017年04期

【摘要】：针对众筹项目由于社会影响力不足而成功率较低的问题,提出了面向众筹平台的社交网络影响力预测方法。该方法基于众筹网站和社交网络的实时观测数据,分别提取累积和增量等多类别预测特征,并随着社会推广的进行而渐进地预测项目的社交网络影响力增益,最后采用带L1一范数约束惩罚的逻辑回归等方法进行预测特征分析。实验结果表明:在整个推广过程中,众筹项目的社交网络影响力可以被精确预测,准确率最高达88.31%;分类器在采用累积特征时具有比采用增量特征更好的预测效果;项目统计特征和推广者的社会影响力等特征具有更高的重要性和更稳定的显著性。该方法成功地解决了众筹项目的社会影响力预测问题,并为设计更好的社交网络推广策略提供了依据。
[Abstract]:Aiming at the problem that the success rate of crowdfunding projects is low due to the lack of social influence, a method for predicting the influence of social networks based on crowdfunding platform is proposed. Based on the real-time observation data of crowdfunding websites and social networks, this method extracts the cumulative and incremental prediction features respectively, and gradually predicts the social network influence gain of the project as the social promotion progresses. Finally, the prediction features are analyzed by using the method of logical regression with L1 norm constraint penalty. The experimental results show that the social network influence of crowdfunding projects can be accurately predicted, and the accuracy is up to 88.31, and the classifier has better prediction effect than incremental feature in the application of cumulative features. The statistical features of the project and the social influence of the promoters have higher importance and more stable significance. The method successfully solves the problem of social impact prediction of crowdfunding projects and provides a basis for designing better social network promotion strategies.
【作者单位】：北京大学信息科学技术学院;伊利诺伊大学芝加哥分校计算机系;北京大学光华管理学院;
【基金】：国家重点基础研究发展计划资助项目(2013CB329600) 国家自然科学基金资助项目(61372191,61572492,61672050)
【分类号】：F724.6;F832;TP393.09

【参考文献】