1 / 1

Mobile Search and Advertisement Cache Architecture

Mobile Search and Advertisement Cache Architecture. Dimitrios Lymberopoulos, Emmanouil Koukoumidis, Jie Liu, Doug Burger (MSR Redmond) Varun Kansal, Chen Xia, Kenny Chien, Bin Wu, Fang Wang, Melissa Dunn (MAXPLAT). Mobile Search Experience. Push Search/Ads to the Phone. Performance Bottleneck.

mayes
Télécharger la présentation

Mobile Search and Advertisement Cache Architecture

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Mobile Search and Advertisement Cache Architecture Dimitrios Lymberopoulos, Emmanouil Koukoumidis, Jie Liu, Doug Burger (MSR Redmond) Varun Kansal, Chen Xia, Kenny Chien, Bin Wu, Fang Wang, Melissa Dunn (MAXPLAT) Mobile Search Experience Push Search/Ads to the Phone Performance Bottleneck Data Sensing Web m.bing.com Data Search/Ad Request Community+ Personal … Data Ads Data 3G indexes Subjective Cache Context + Personalization results Phone Cloud • 3G link is slow • Average end-to-end delay: 4-8 seconds • 3G connection time: 10 - 30 seconds • Challenges • What information do we cache? • How do we cache it? • How do we manage it over time? • Leverage mobile trends • Increasing flash density at lower cost • Increasing processing power Goal: Faster Mobile Search Experience SONGO (Search ON the GO): A Data-driven Architecture Mobile Search Log Analysis - 100M Queries Fetch links Results Page Construction results Ranking Rank links based on their quality user clicks Link Storage (Flash ~ 1-2MB) cache hit results cache hit Personalization query Community Aggregate Volume (%) SONGO cache query Percentage of Unique Users Phone periodic updates cache miss m.bing.com Hash Table (RAM ~ 200KB) Adjust Ranking Scores SONGO cache Up-to-date Community Cache 60% of mobile query volume hits 6K queries and 4K links! 50% of the users repeat a query at least 70% of the time! … … user clicks Number of query-link pairs Probability of a new query-link pair 0.links 15.links 31.links Experimental Prototypes Beyond Web Search On average 66% of the queries a user submits hit the cache • Real-time: • search results • business lookups • ad delivery • Opportunities: • Faster user experience • Monetization of autosuggest • Fastest mobile ad delivery engine • Personalized ranking for search/ads • Privacy: profile on the phone 23x more energy efficient! 16x faster!

More Related