Search engine advertising has become the main stream of online advertising system. In current search engine advertising systems, all users will get the same advertisement rank if they use the same query. However, different users may have different degree of interest to each advertisement even though they query the same word. In other words, users prefer to click the interested ad by themselves. For this reason, it is important to be able to accurately estimate the interests of individual users and schedule the advertisements with respect to individual users' favorites. For users that have rich history queries, their interests can be evaluated using their query logs. For new users, interests are calculated by summarizing the interests of other users who use similar queries. In this paper, we provide a model to automatically learn individual user's interests based on features of user history queries, user history views of advertisements, user history clicks of advertisements. Then, advertisement schedule is performed according to individual user's interests in order to raise the clickthrough rate of search engine advertisements in response to each user's query. We simulate user's interests of ads and clicks in our experiments. As a result, our personalized ranking scheme of delivering online ads can increase both search engine revenues and users' satisfactions.
[1]
Feng Qiu,et al.
Automatic identification of user interest for personalized search
,
2006,
WWW '06.
[2]
Anindya Ghose,et al.
An empirical analysis of sponsored search performance in search engine advertising
,
2008,
WSDM '08.
[3]
Benjamin Piwowarski,et al.
Predictive user click models based on click-through history
,
2007,
CIKM '07.
[4]
Abdur Chowdhury,et al.
A picture of search
,
2006,
InfoScale '06.
[5]
John Tomlin,et al.
Optimal delivery of sponsored search advertisements subject to budget constraints
,
2007,
EC '07.
[6]
Aranyak Mehta,et al.
AdWords and Generalized On-line Matching
,
2005,
FOCS.
[7]
Amin Saberi,et al.
Allocating online advertisement space with unreliable estimates
,
2007,
EC '07.
[8]
Brenda L. Dietrich,et al.
A Column Generation Approach for Combinatorial Auctions
,
2002
.
[9]
Matthew Richardson,et al.
Predicting clicks: estimating the click-through rate for new ads
,
2007,
WWW '07.
[10]
S. Sathiya Keerthi,et al.
Constructing a maximum utility slate of on-line advertisements
,
2007,
ArXiv.
[11]
Ashish Goel,et al.
Truthful auctions for pricing search keywords
,
2006,
EC '06.