## Online Search and Advertising, Future and Present

author: Chris Burges, Microsoft Research
published: Dec. 20, 2008,   recorded: December 2008,   views: 449
Categories
You might be experiencing some problems with Your Video player.

# Slides

0:00 Slides Online Search and Advertising, Future and Present Contents ~ Search and Advertising ~ Why Search Works… Key Points What‟s wrong with what we do now? How might ads be targeted better? User Models Key Points What About Search? Search: Somewhere in the Near Future Example : MS LiveSearch Search: Somewhere in the Near Future How to get the information we need, to build good models for users? Key Points Search Applications: And,Data Changes Everything (1) Search Applications: And,Data Changes Everything (2) Data Changes Everything (1) Data Changes Everything (2) Key Points Key Points How To Proceed? The Eliza Effect Our Prime Directive in Building Sam: Let the Data do the Work Using Category Graphs to Drive Dialog Use Category Graphs to Build Models Other Useful Sources of Data Temporal Querying Behavior We Are Not Alone One Possible Sentence Generator New Challenges for Machine Learning Demo ~ Some New Results on Ranking ~ Empirical Optimality of lambda -rank Some IR Measures IR Measures, cont. LambdaRank: Background The RankNet Cost (1) The RankNet Cost (2) RankNet Cost ~ Pairwise Cost Pairwise Cost Revisited LambdaRank The Lambda Function Lambda Functions for MAP, MRR Local Optimality Data Sets Which lambda-function to choose? (1) Which lambda-function to choose? (2) Sampe size matters IR Measure Optimality -Conclusions ~ RSA, Factoring, and Optimization ~ Factoring biprimes as optimization Circumstantial Evidence That Factoring is Not NP-hard Is This The Best We Can Do? RSA Challenge Represent the Problem in Binary First Trick: Linearization Linearization, cont. A Geometrical Problem (1) Second Trick: Quantization (2) More Simple Tricks The Geometric View Projections Lose Information Conclusions - questions

# Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Lecture popularity: You need to login to cast your vote.

# Description

Search engine companies are gathering treasure troves of user-generated data. It has already been shown that such data can be used to directly improve the user's online experience. I will discuss some ideas as to what online search and advertising might look like a few years hence, in light of the algorithms and data we have now. Moving from future to present, I will outline some recent work done by researchers in the Text Mining, Search and Navigation team at Microsoft Research; the work in TMSN touches many aspects of online search and advertising.