行为定向广告作为一种新的精准营销手段在近几年逐渐兴起,这种广告模式以其及时、精准、高效的特点备受人们的关注。行为定向广告的主旨是利用用户的网络浏览行为,提供符合用户意图的广告信息,它主要通过分析用户的网页历史访问记录来挖掘有价值的用户行为信息,并针对这种信息投放与之相关的广告。针对行为定向广告问题,提出了一种全新的行为定向广告投放算法,该算法首先根据用户行为特征模型对用户最近访问的网页按主题进行聚类,然后利用用户行为特征分析算法对每一类网页进行行为特征分析并计算该类网页的权重,利用该权重以及该类网页的质心与广告的相似度来计算最后得分,并按照这个得分对广告进行排序从而选出适合该类网页的广告。大量实验表明这是一种高效的行为广告投放算法,有着十分广阔的应用前景。
Behaviour-targeted advertising as a new means of precision marketing is rising in recent years,it's attractive to people for its characteristics of in time,accuracy and efficiency.The purport of this method is to provide advertising information in compliance with users' intention by employing their internet browsing behaviours.It mines valuable user behaviour information mainly through analysing historical webpages access records of the users and then delivers correlated advertisements pertinent to the information.This paper presents a novel algorithm with regard to behaviour-targeted advertising in light of the issue of behaviour-targeted advertisement.First,the algorithm clusters on themes the webpages which users recently visited according to user behaviour feature model,and then uses user behaviour feature analysis algorithm to analyse the behaviour features on webpages of each category and to calculate the weight of the webpage of such category.By using the weights above and the similarity between the centroid of webpage of each category and the advertisements,the method calculates final scores,with these scores the advertisements are ranked and the ones suiting the webpage of a category are selected.A large number of experiments show that it is a highly efficient advertising algorithm with very broad application prospects.