微博中基于多路径目标的广告投送技术  

Online advertising technique based on desired set of users via different paths in microblogs

作　　者：何永强[1] 秦勤[1] 王俊鹏[2] 

机构地区：[1]河南工程学院计算机学院,河南郑州450007 [2]北京理工大学计算机学院,北京100081

出　　处：《计算机工程与设计》2016年第10期2733-2737,共5页Computer Engineering and Design

基　　金：河南省高等学校重点科研基金项目(15A520054);河南省科技厅科技计划课题基金项目(112102310550)

摘　　要：基于微博广告平台,提出两个方法增加广告营销的准确性。第一是识别近似话题,给定一个指定的话题查询,在微博中近似地识别同样的听众,其思想是代替初始的原始话题,对近似话题出价,用尽可能小的成本,使广告达到同样的用户数目;提出一个算法,根据用户指定的话题查询识别专家,根据专家经验对专家分类。该方法在精确营销中起到重要作用。利用大规模的Twitter实验数据评估所提算法,实验验证了算法的精确性和有效性。Two problems were introduced to marketing campaigns and aid ad on Twitter advertising platforms.Analogous topics were identified where the same audience was found approximately for a given query topic on the Twitter platform.The reason is that the same audience was reached approximately by bidding on an analogous topic instead of the original query topic,while spending less of budget.Algorithms were presented to identify expert users on a given query topic.A large dataset from Twitter was collected for evaluating the proposed algorithms,and results show that the efficiency and accuracy were higher than alternative approaches.

关 键 词：类似话题 专家分类 广告营销 微博算法 话题分类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]