Efficient Behavior Targeting Using SVM Ensemble Indexing Jun Li, Peng Zhang, Yanan Cao, Ping Liu, Li Guo Chinese Academy of Sciences State Grid Energy Institute, China Behavior targeting Behavior Targeting (BT) uses users’ historical behavior data to select the most relevant ads for display. ads User behavior data Example from Yahoo! Research Targeted users Regression for BT Poisson Regression model (Ye Chen, eBay, 2009). x: ad clicks and views, page views, search queries and clicks. y: click-through rate (CTR). Poisson dis. Poisson reg. on view View data ad category Click data Poisson dis. Poisson reg. on click Ye Chen et al., Large-scale behavior targeting (KDD’09 best paper award) Limitations Limitations: parameter tuning is very difficult. the Poisson assumption is not always true for real-world behavior data. Clicks are typically several orders of magnitude fewer than views. User interests are not always fixed, but rather transient. Classification for BT SVM for classification View data ad category Click data View and click data(+) View but no click data(-) SVM for classification Challenges 1,2,3 Example 1: 3 users on Nikon (www.nikon.com)’s ad a Classification for BT Ensemble SVM on data streams Merits no complicated parameters no statistical assumptions Dynamic model on data streams Challenge 4 Limitations Time cost is heavy for online computing ensemble prediction time cost: A (advertisers)*W(ensemble size)*N(support vectors)*T(features) Example 2: We collect 2 million behavior events (W = 10) in 1 minute, and prediction result costs 53 minutes. Solutions Construct Index structure for Ensemble SVM. Features map Text terms Support vector Document Ensemble SVM Document set Why the index work ? Trade space for time. shared features among multiple support vectors Zhang et al., of knowledge index for online data streams the sparseP. structure support vectors ( KDD 2011 & ICDM 2011) The index structure The SVM-index structure Example 3: based on example 1, consider a SVM with 3 support vectors Inverted hashing table Support vectors Ensemble information Time complexity O(T) The index structure Operations – Search: Predict the label of each incoming user data x, • Step 1: searches support vectors in the left inverted indexes • Step 2: calculate x’s class label See our source codes. – – Insert: Integrate new classifiers into ensemble Delete: Drop outdated classifiers from ensemble Memory Experiments Data sets Search • engine data Comparisons – – – Possion E-SVM E-Index (our method) Comparisons Observations E-index has sub-linear prediction time E-SVM consumes more memory Comparisons Ensemble models are more accurate than Poisson regression model Comparisons The index method can significantly improve the efficiency, especially when the ensemble size is large. Related Work Behavior targeting Regression models vs. classification models Stream indexing Boolean expression indexing in Publish/subscribe systems Ensemble models Concept drifting Conclusions Contributions Identify and address the prediction efficiency problem for ensemble models for behavior targeting. Convert ensemble SVM model to a document set, and propose a new type of invert text index structure to achieve sub-linear prediction time. Future work Index more complicated SVM models with non-linear kernels. Questions? For source code, visit our website streamming.org/homepages/lijun.html