Advertising Sciences is an inter-disciplinary field that studies the dynamics of an ecosystem of users, publishers, advertisers, and ad networks. Its central problem is to find the “best” matching ads to a user in a given context (e.g., query, page view) that optimize the utilities of the participants in the ecosystem under certain business constraints (blocking, targeting, guaranteed delivery, etc) by applying cutting-edge algorithms and techniques in information retrieval, machine learning, data mining, optimization, and micro-economics.
Featured Project
Team
- Ajay Shekhawat
- Alyssa Glass
- C. Nagarajan
- Chris Bartels
- David Pardoe
- Dinesh Garg
- Dongni Chen
- Dragomir Yankov
- Gagan Bansal
- Haibin Cheng
- H. Bommaganti
- Jason Zien
- Jignesh Parmar
- Jinhui Liu
- Kannan Achan
- Kara Manatt
- Karim Filali
- Ken Mallon
- Kin Fai Kan
- Krishna P. Leela
- Kun Liu
- Lei Tang
- Leo Neumeyer
- Maurício Mediano
- Nagaraj Kota
- N. L. Bhamidipati
- Patrick R Jordan
- Peiji Chen
- P. Krishnamurthy
- R. Zhang
- Rushi Bhatt
- Sachin Garg
- S. Kshetramade
- Sharath Rao
- S. Mukherjee
- S. Bhattacharya
- Stefan Schroedl
- Sunil Jagadish
- Tasos Anastasakos
- V. Shashikant Chaoji
- Wanlin Pang
- Wei Li
- Wei Ye
- Yang Zhou
- Yayati Kasralikar
- Ying Cui
- Zhen Guo
Projects
S4: Distributed Stream Computing PlatformS4 is a real-time MapReduce software platform that is used to process massive streams of data. The information is used to improve search and advertising experience by providing fresh and personalized results to consumers and advertisers.
Read more
Ad Indexing & Retrieval Displaying ads alongside web queries is a very effective advertising approach. YLabs is working on better ways to index large advertiser databases and performing efficient and precision driven real-time retrieval for serving sponsored search ads.
Read more
Estimation of Reserve Price for Sponsored Search Auctions Reserve Pricing is an important feature of an auction marketplace. offers several benefits such as offering price support and improving the user experience by eliminating ads with poor relevance
Read more
Machine Learned Categorization for Ads and QueriesMachine Learned Categorization efforts at Y! labs focus research and development of automated methods of categorizing the key entities in advertising - users, content, queries, and ads to facilitate superior matching and improved relevance.
Read more
Predicting Query to Ad Relevance Sponsored search needs to satisfy both the search users, by providing high quality advertisements that are relevant to the user, and the advertiser, by driving customers with a buying intent to their site.
Read more
Display Supply & Demand ForecastingY! Labs is developing the state-of-the-art forecasting algorithms to predict future supply of target-able page-views and demand from advertisers
Read more
Mapping Search Query Language to Advertiser Bidded Terms Sponsored search is aimed at connecting search advertisers to people who may be interested in their products, but advertisers and consumers don't always speak the same language.
Read more
Contextual Ads Relevance Modeling Y! Labs is developing advanced information retrieval techniques to improve contextual ads relevance.
Read more
Contextual Advertising Y! Labs is pioneering advances in contextual advertising by building highly scalable ad search systems and predictive models for effectively targeting ads based on the context of the user's browsing behavior and the content of the web pages visited.
Read more
Response Prediction Y! Labs is developing advanced machine learning and statistical modeling techniques to predict user response (click & conversion) to ads impressions given user context.
Read more
Conversion Modeling in Sponsored Search Both advertisers and search users want to increase conversion rates and decrease the cost. This project measures and predicts conversion rates and uses this data to influence ad ranking, pricing and placement, for better value for both search users and advertisers.
Read more
Traffic Quality Measurement & PricingYahoo! Labs is working on a methodology drawing from advanced statistical estimation techniques to measure traffic quality for a heterogeneous mix of publishers.
Read more
Personalized Ad Placement in Web SearchWe are working on models that give personalized predictions of the response to sponsored search ads. These models will improve the user experience by making changes to ranking and presentation of ads, according to the user's past behavior.
Read more
Generalized Utility in Sponsored Search Auctions Yahoo! Labs is constantly improving Yahoo!’s ad auction system to create long-term value to search engine users, publishers, and advertisers
Read more
Display Inventory Allocation Optimization Inventory allocation plays a critical role in revenue management for the online advertising industry by supporting both Admission Control and Ad Serving.
Read more
KeystoneYahoo! Labs is working on the next generation of contextual advertising technology.
Read more
Internationalization of Click Models Yahoo! Labs is creating a global platform leveraging highly parallel infrastructure to process and train billions of samples to create click prediction models tailored for different global markets.
Read more
User TargetingYahoo! Labs is building a state-of-the-art targeting system that brings the highest quality traffic to advertisers and optimizes revenue for publishers by serving the most relevant ads to users.
Read more
Semantic AnalysisY! Labs is working on inferring user's intent from the content of the page that the user is viewing, in order to provide effective contextual advertising.
Read more