1. Focused Matrix Factorization for Audience Selection in Display Advertising BHARGAV KANAGAL, AMR AHMED, SANDEEP PANDEY, VANJA JOSIFOVSKI, LLUIS GARCIA-PUEYO, JEFF YUAN PRESENTER: I GDE DHARMA NUGRAHA CHONNAM NATIONAL UNIVERSITY

2. Outlined  Introduction  Problem in Matrix Factorization  Focused Matrix Factorization Model  Model Learning and Inference  Implementation  Experimental Evaluation  Conclusion

3. Introduction  Audience selection or audience retrieval is the problem in display advertising to display ads for those users who are most likely to show interest and respond positively to the campaigns.  The user’s past feedback on this campaign can be leveraged to construct such a list using collaborative filtering techniques such as matrix factorization.  However, the user-campaign interaction is typically extremely sparse, hence the conventional matrix factorization does not perform well.  Moreover, simply combining the users feedback from all campaigns does not address this since it dilutes the focus on target campaign in consideration.

4. Introduction  To resolve these issues, this paper propose a novel focused matrix factorization (FMF) which ◦Learns users’ preference towards the specific campaign products, while also exploiting the information about related products. ◦Exploit the product taxonomy to discover related campaigns and design models to discriminate between the users’ interest towards campaign products and non-campaign product.

5. Introduction  The illustration of different approach in this paper is shown in the figure.

6. Problem in Matrix Factorization

12. Focused Matrix Factorization Model

19. Model Learning and Inference

21.  To use SGD, a term from the summation is sampled, which denote using (i, u 1, u 2 ). Depending on whether the item is from the target campaign (i.e., i ϵ T) or from some non-target campaign j (i.e., i ϵ N j ), this model obtain two sets of gradients which are show in the figure.

22. Implementation  This model is developed using C++ for a multi-core implementation and BOOST library package for storing the factor matrices.  The global state maintained by the SGD algorithm consists of the 3 factor matrices {v S, v N, v I } and the α vector. A lock is introduced for each row in factor matrices.  In the SGD algorithm, in each iteration of training, its execute 3 steps. ◦The first step, sampling a 3-tuple (i, u 1, u 2 ). ◦The second step, read the appropriate user and item factors and compute the gradients with respect to them. ◦The third step, update the factor matrices based on the gradients.  Using locks over such small vector can result in significant increase in the processing time. To alleviate this problem, caching technique is proposed.

23. Implementation

24. Experimental Evaluation  Experimental Setup ◦Dataset for evaluation use the log of previous advertising campaigns obtained from a major advertising network. ◦The dataset contains information about the item corresponding to various advertising campaigns and an anonymized list of users who actually responded to the campaign by making a purchase of the campaign item. ◦In addition, the dataset that contain a taxonomy over the various item in the campaign. ◦It contain 50.000 users and around a million items in the taxonomy. ◦The taxonomy dataset contains 3 level deep, with around 1500 nodes at lowest level, 270 at the middle level and 23 top level categories. ◦Overall, the dataset contain 23 campaigns.

25. Experimental Evaluation

26.  Cross-validation/Parameter Sweep ◦For each of the experiments, a parameter sweep over MapReduce cluster was executed. ◦The parameter that sweep over included U,  I, N and K, the number of factors. ◦For each setting of parameter was evaluated in 4 different initializations and picked the best initialization for each configuration, in terms of performance on the validation dataset. ◦AUC was choose over the test set for a given number of factors to report the experiments.

27. Experimental Evaluation  Experimental Results ◦The first experiment, GMF, MF and FMF2 technique was compared for different campaigns. ◦The figure show the result

28. Experimental Evaluation  Experimental Results ◦The second experiment, performance over the individual campaigns was examined. ◦The figure show the result

29. Experimental Evaluation  Experimental Results ◦The third experiment, to examine the best performance model across all factor sizes for each campaign. ◦The figure show the result

30. Experimental Evaluation  Experimental Results ◦The forth experiment, to examine the influence of taxonomy for each model. ◦The figure show the result.

31. Experimental Evaluation  Experimental Results ◦The last experiment, to compare the performance between FMF model. ◦For each of the four campaigns, FMF1, FMF2 and FMF3 model was trained. ◦For the figure, the models FMF1 and FMF2 perform much better than FMF3. ◦The reason is caused by FMF3 have much more constrained than the other two models.

32. Experimental Evaluation  Experimental Results ◦Effect of Campaign Size ◦The figure show the result for different campaign size ◦The performance of FMF2 model is a function of the target campaign size, i.e., the number of items in the target campaign. ◦From the figure, show that the performance of FMF2 is robust and largely unaffected by the campaign size.

33. Experimental Evaluation  Experimental Results ◦Effect of Intra-campaign relationship (Campaign Homogeneity) ◦In this experiment, the performance of FMF2 models as a function of the homogeneity of the target campaign was explored. ◦From the figure show that the AUC scores increase as long as the homogeneity of the campaign.

34. Experimental Evaluation  Experimental Results ◦Effect of Inter-campaign relationship (Information Transfer) ◦This experiment explores the effect of inter-campaign relationship for information transfer in the FMF2 model. ◦This experiment pick a fairly homogeneous campaign X and split it into two parts X 1 and X 2. Then picked another campaign Y and constructed two configuration using X 1, X 2 and Y. X 1 become the target campaign, in config 1, X 2 become the non-target campaign and in config 2, Y become the non- target campaign. ◦The figure show that config 1 has higher AUC score than config 2 since config 1 has X 2 as the non-target campaign which is highly similar to X 1.

35. Experimental Evaluation  Experimental Results ◦The last experiment, to compare the performance between FMF model. ◦Efficiency ◦In this experiment, the trade-offs that is obtained by using the caching technique is demonstrated. ◦The result show in the figure. When the threshold is set to 0, there is complete synchronization. As the threshold is increased, the synchronization with the global copy is performed less often, resulting in faster runtime but less accuracy.

36. Conclusion  This paper propose Focused Matrix Factorization (FMF) model to appropriately borrow relevant information from other campaign while still retaining focus on the target campaigns.  The experiment result show that FMF model consistently outperforms the traditional matrix factorization techniques over all kinds of campaigns.  In addition, the experiment resulting the character of the conditions which the approach will obtain significant improvements.