許永真臺灣大學：資訊工程學研究所林光龍Lin, Koung-LungKoung-LungLin2007-11-262018-07-052007-11-262018-07-052005http://ntur.lib.ntu.edu.tw//handle/246246/54083Recommendation research has achieved successful results in many application areas. However, for supermarkets, since the transaction data is extremely skewed in the sense that a large portion of sales is concentrated in a small number of best selling items, collaborative filtering based customer-triggered recommenders usually recommend hot sellers while rarely recommend cold sellers. But recommenders are supposed to provide better campaigns for cold sellers to increase sales. In this thesis, we propose an alternative ``item-triggered' recommendation to identify potential customers for cold sellers. In item-triggered recommendation, the recommender system will return a ranked list of customers who are willing to buy a given item. This problem can be formulated as a problem of classifier learning, but due to the skewed distribution of the transaction data, we need to solve the rare class problem, where the number of negative examples is much larger than the positive ones. We present a boosting algorithm to train an ensemble of SVM classifiers to solve the rare class problem and compare the algorithm with its variants. We apply our algorithm to a real-world supermarket database and use the area under the ROC curve (AUC) metric to evaluate the quality of the output ranked lists. Experimental results show that our algorithm can improve from a baseline approach by about twenty-three percent in terms of the AUC metric for cold sellers which is as low as 0.64\% of customers have ever purchased.1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.3.1 Bayesian Optimal Recommendation . . . . . . . . . . . . 14 1.3.2 Types of Recommender . . . . . . . . . . . . . . . . . . 17 1.4 Outline of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . 21 2 Related Work 22 2.1 Recommender System . . . . . . . . . . . . . . . . . . . . . . . 22 2.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.1.2 Taxonomy of Recommendation Techniques . . . . . . . . 24 2.1.3 Content-Based Recommendation . . . . . . . . . . . . . . 27 2.1.4 Collaborative Filtering Based Recommendation . . . . . . 28 2.2 Support Vector Machine . . . . . . . . . . . . . . . . . . . . . . 30 2.3 Rare Class . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 2.4 Boosting Algorithms . . . . . . . . . . . . . . . . . . . . . . . . 34 3 Item-triggered Recommendation 42 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 3.2 System Framework . . . . . . . . . . . . . . . . . . . . . . . . . 43 3.3 Boosting SVM . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.4 Varation of Boosting SVM . . . . . . . . . . . . . . . . . . . . . 52 3.5 From item-triggered to customer-triggered . . . . . . . . . . . . . 54 4 Experiments 55 4.1 Ta-Feng Data Set . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.2 Cold Sellers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.3 Experiment Environment . . . . . . . . . . . . . . . . . . . . . . 63 4.4 Evaluation Methodology . . . . . . . . . . . . . . . . . . . . . . 64 4.5 Potential Customer of Cold Seller Detection . . . . . . . . . . . . 66 4.6 Targeted Advertisement of Unsought Products . . . . . . . . . . . 69 4.7 Cold Sellers Recommendation . . . . . . . . . . . . . . . . . . . 74 4.7.1 Removing Hot Sellers from Data . . . . . . . . . . . . . . 75 4.7.2 Only Recommend Cold Sellers . . . . . . . . . . . . . . . 80 5 Conclusion and FeatureWork 845110554 bytesapplication/pdfen-US推薦系統支撐向量機普適提演算法稀少類別分類商品驅動消費者驅動未請求商品recommender systemsupport vector machineboosting algorithmrare class classificationitem-triggeredcustomer-triggeredunsought productthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/54083/1/ntu-94-D87526007-1.pdf