摘要
基于哔哩哔哩网站的UP主数据,建立内容创作者的商业价值评估模型。该模型以UP主的商业广告报价为预测目标,体现出较强的泛化能力和良好的实用性。使用熵权法对内容创作者的八项特征加权,汇集成曝光度、认同度、讨论度3项指标,然后使用多层感知机回归算法对广告报价进行预测。结果表明,模型能够较为准确地预测UP主广告报价,对内容创作者和广告商等群体具有一定的参考价值。
A commercial value assessment model for video creators is established based on the dataset of up-loaders of a video website named Bilibili.The model takes the advertising price of up-loaders as the prediction target,which shows strong generalization ability and good practicality.The entropy weighting method is used to weigh eight features of video creators,which are gathered into three indicators,namely exposure,recognition,and discussion.Then we use the multi-layer perceptron(MLP)regression algorithm to predict the advertising price.The results show that the model can predict up-loaders’advertising prices accurately and help advertisers make business decisions.
作者
吕慧敏
李昕
LV Hui-min;LI Xin(School of Electronics and Information Engineering,Liaoning University of Technology,Jinzhou 121001,China)
出处
《辽宁工业大学学报（自然科学版）》
2024年第3期164-168,共5页
Journal of Liaoning University of Technology(Natural Science Edition)
关键词
B站
熵权法
多层感知机回归
商业价值评估
Bilibili
entropy weight method
MLP regression
commercial value assessment
