[1]郝晓培,朱建生,单杏花.基于出行关系的广告点击率预测模型的研究[J].计算机技术与发展,2022,32(05):136-140.[doi:10. 3969 / j. issn. 1673-629X. 2022. 05. 023]
　HAO Xiao-pei,ZHU Jian-sheng,SHAN Xing-hua.Research on Forecast Model of Advertisement Click Rate Based on Travel Relationship[J].,2022,32(05):136-140.[doi:10. 3969 / j. issn. 1673-629X. 2022. 05. 023]
点击复制

基于出行关系的广告点击率预测模型的研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年05期
页码:
136-140
栏目:
应用前沿与综合
出版日期:
2022-05-10

文章信息/Info

Title:
Research on Forecast Model of Advertisement Click Rate Based on Travel Relationship
文章编号:
1673-629X(2022)05-0136-05
作者:
郝晓培; 朱建生; 单杏花
中国铁道科学研究院,北京 100081
Author(s):
HAO Xiao-pei; ZHU Jian-sheng; SHAN Xing-hua
China Academy of Railway Sciences,Beijing 100081,China
关键词:
广告投放; 个性化推荐; 图注意力网络; Wide&; Deep 模型; 关系网络
Keywords:
advertising; personalized recommendation; graph attention network / GAT; Wide&; Deep model; network of relationships
分类号:
TP301. 6
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 05. 023
摘要:
随着智能移动终端的普及以及互联网产品的多样化发展,在为用户提供服务的同时也丰富了广告的投放渠道,然而传统的广告平台投放策略是针对全体用户批量投送,严重影响用户体验,同时也降低了点击率以及转化率,为广告平台的可持续发展带了极大的挑战,广告的精准投放已经成为互联网服务产品领域内的研究热点之一。 该文依托铁路 12306互联网售票系统的广告平台,在图注意力网络 GAT 与 Wide&Deep 模型的基础上提出了一种新型的广告个性化推荐模型。该模型首先将用户的同行的关系、购票关系以及用户与广告的交互关系作为基础数据构建关系网络,利用图神经网络聚合邻节点特征以及自身节点特征以实现在非欧氏空间新节点的表示向量的更新,生成最终的特征向量并作为 Wide&Deep模型的输入实现广告点击率预测。 论文利用近半年的广告平台数据对该模型的性能进行评估,实验效果显示,该模型能够准确对广告点击率进行预测,实现了广告的精准投放。
Abstract:
With the popularity of intelligent mobile terminals and the diversified development of Internet products,it not only provides services for users,? ? ? but also enriches the advertising channels. However,the traditional advertising platform delivery strategy is aimed at all users,which seriously affects the user experience and reduces the click through rate and conversion rate,bringing great challenges to the sustainable development? ? ? ?of the advertising platform. Accurate advertising has become one of the research hotspots in the field of Internet service products. Relying on? ?the advertising platform of railway 12306 Internet ticketing system, we propose a new personalized advertising recommendation model based? ?on the graph attention network gat and wide & deep model. The model first constructs the relationship network by taking the user’s peer relationship, ticket purchasing relationship and the interaction between user and advertisement as the basic data. The graph neural network is used to aggregate the features of adjacent nodes and its own nodes to update the representation vector of new nodes in non Euclidean space,and the final feature vector is generated and used as the input of wide & deep model to realize the advertising click through rate prediction. In this paper,the performance of the model is evaluated by using the advertising platform data of nearly half a year. The experimental results show that the model can accurately predict the click through rate of advertising,and achieve the accurate delivery of advertising.

相似文献/References:

[1]查文琴 梁昌勇 曹镭.基于用户聚类的协同过滤推荐方法[J].计算机技术与发展,2009,(06):69.
　ZHA Wen-qin,LIANG Chang-yong,CAO Lei.Collaborative Filtering Recommendation Method Based on Clustering of Users[J].,2009,(05):69.
[2]曹毅 贺卫红.基于内容过滤的电子商务推荐系统研究[J].计算机技术与发展,2009,(06):182.
　CAO Yi,HE Wei-hong.Research on E- Commerce Recommender System Based on Content - Based Filtering[J].,2009,(05):182.
[3]汤亚玲 秦峰.Web行为下的正向关联规则挖掘研究[J].计算机技术与发展,2007,(08):40.
　TANG Ya-ling,QIN Feng.Research of Forward Association Rules Mining Under Web Behaviour[J].,2007,(05):40.
[4]卫琳.基于搜索结果的个性化推荐系统研究[J].计算机技术与发展,2007,(09):65.
　WEI Lin.A Study of Personalization Recommendation System Based on Search Result[J].,2007,(05):65.
[5]但微 才书训.电子商务中Web挖掘技术的应用探讨[J].计算机技术与发展,2006,(01):207.
　DAN Wei,CAI Shu-xun.Using Web Mining in Electronic Commerce[J].,2006,(05):207.
[6]王嫣然 陈梅 王翰虎 张鑫.一种基于内容过滤的科技文献推荐算法[J].计算机技术与发展,2011,(02):66.
　WANG Yan-ran,CHEN Mei,WANG Han-hu,et al.A Content-Based Filtering Algorithm for Scientific Literature Recommendation[J].,2011,(05):66.
[7]杨晶,成卫青,郭常忠.基于标准标签的用户兴趣模型研究[J].计算机技术与发展,2013,(10):208.
　YANG Jing[],CHENG Wei-qing[],GUO Chang-zhong[].Research on User Interest Model Based on Standard Tag[J].,2013,(05):208.
[8]殷凤霞.社会网络中基于内容语义的新闻推荐方法研究[J].计算机技术与发展,2013,(10):253.
　YIN Feng-xia.Research on Method of News Recommendation Based on Content Semantic in Social Network[J].,2013,(05):253.
[9]黄谭,苏一丹.基于混合用户模型的二分图推荐算法[J].计算机技术与发展,2014,24(06):145.
　HUANG Tan,SU Yi-dan.Bipartite Graph Recommendation Algorithm Based on Hybrid User Model[J].,2014,24(05):145.
[10]吴楠,秦锋,姜太平. 一种基于用户情境聚类的个性化推荐算法[J].计算机技术与发展,2014,24(10):106.
　WU Nan,QIN Feng,JIANG Tai-ping. A Personalized Recommendation Algorithm Based on User Context Clustering[J].,2014,24(05):106.

更新日期/Last Update: 2022-05-10