簡易檢索 / 詳目顯示

回結果列表
研究生： 程柏勳
Cheng, Bo-Syun
論文名稱： 發掘主題多樣性之個人化推薦系統
Discovering Topic Diversity in Personalized Recommendation System
指導教授： 蔣榮先
Chiang, Jung-Hsien
學位類別： 碩士
Master
系所名稱： 電機資訊學院 - 資訊工程學系
Department of Computer Science and Information Engineering
論文出版年： 2019
畢業學年度： 107
語文別： 英文
論文頁數： 46
中文關鍵詞： 推薦系統 、主題多樣性 、協同式過濾 、推薦可解釋性 、多類別分類 、遷移學習
外文關鍵詞： Recommender System, Topic Diversity, Collaborative Filtering, Explainable Recommendation, Multi-label Classification, Transfer Learning
相關次數： 點閱：109  下載：8 
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報

    • 隨著網路資訊爆炸性地增長，導致使用者無法輕易地從海量的資訊中挑選自己的需求，因此推薦系統在我們的生活中扮演著重要的角色。為了提供更合適的推薦服務給使用者，近年來不少研究致力於探索使用者的喜好來發展個人化的推薦系統。其中，基於協同式過濾的推薦系統根據不同使用者與商品之間的互動紀錄，發掘使用者潛在的喜好。然而該方法往往會忽略使用者對於商品本身所具備特性的喜好，且無法解釋推薦的結果。
    本研究的目的是建立發掘使用者在商品主題上喜好的機制來強化推薦系統。我們根據使用者過往互動過的商品主題資訊來發掘使用者在主題性上的喜好，進而使其推薦符合使用者喜好的商品，並且透過分析使用者在主題性上的喜好分布與商品本身的主題性來提供使用者可理解的推薦原因。
    在本研究採用了兩個實際的資料集進行推薦結果的評估實驗，實驗結果說明了本研究提出的方法不僅比傳統的方法有更好的推薦表現，並證明引入使用者在主題性上的喜好分布的確會使原本的推薦系統達到更好的效果。並且，本研究提出數個案例探討來說明使用者在主題性的喜好如何影響推薦結果，我們發現引入使用者在主題上的喜好後推薦的商品確實更符合使用者的喜好。最後，期望本研究能應用於現實生活中來提升使用者的購物體驗。

    Due to the information explosion on the internet, people cannot easily find what they really want. Therefore, recommendation systems play an important role to our life. To provide more suitable recommendation service to the user, there are several research are committed to building personalized recommendation system by discovering user preference. Among of them, collaborative filtering is modeling user preference on the items according to their past interaction. However, it ignores the user preference on the attribute of the item and the recommendation result is unexplainable.
    Our research aims to build a topic diversity detection model to discover user preference on the topics to enhance the recommendation model. We discover user preference according to the information of item which is viewed by the user in the past, then the recommendation model can recommend proper items which is matched the user preference on the topic. Furthermore, with the user preference on the topics, we can explain that the recommendation list which is based on the user preference.
    We conduct our experiments on the two real-world dataset and evaluate the recommendation performance. The experimental results demonstrate that our proposed model outperforms the traditional methods and shows the effectiveness of the topic diversity. Moreover, we provide several case studies to show how the topic diversity detection model influence the recommended result and the explainability of our model. Finally, we hope that our personalized recommendation system with topic diversity can be applied to real-world recommendation scenario and improve the user’s buying experience.

    中文摘要 I Abstract III 誌謝 V Contents VI List of Tables VIII List of Figures IX Chapter 1 Introduction 1 1.1 Background and Motivation 1 1.2 Research Objectives 2 1.3 Thesis Organization 3 Chapter 2 Related Work 5 2.1 Recommendation System 5 2.1.1 Collaborative Filtering Approach 5 2.1.2 Content-Based Approach 7 2.1.3 Session-Based Recommendation System 8 2.2 Recommendation System with Auxiliary Information 9 Chapter 3 User Preference on Topic Diversity 11 3.1 Overview 11 3.2 Multiple Representations on User Preference 12 3.3 Topic Diversity Detection by Supervised Learning 13 Chapter 4 Recommendation System with User Preference on Topic Diversity 18 4.1 Pairwise Learning Method with User Weight 18 4.2 Joint Learning 20 4.3 Session-Based Recommendation with Topic Diversity 21 Chapter 5 Experimental Design and Results 24 5.1 Dataset and Preprocessing 24 5.2 Experimental Design 25 5.2.1 Performance Comparison 25 5.2.2 Hyper Parameter Influence of Model Performance 27 5.3 Experimental Results 28 5.3.1 Performance Comparison 28 5.3.2 Hyper Parameter Influence of Model Performance 29 5.3.3 Effectiveness of Topic Diversity 31 5.4 Case Study and Discussion 31 Chapter 6 Conclusion and Future Work 40 Reference 42

    [1] B.Sarwar, G.Karypis, J.Konstan, andJ.Reidl, “Item-based collaborative filtering recommendation algorithms,” in Proceedings of the tenth international conference on World Wide Web - WWW ’01, 2001, pp. 285–295.
    [2] Y.Koren, R.Bell, andC.Volinsky, “Matrix Factorization Techniques for Recommender Systems,” Computer (Long. Beach. Calif)., vol. 42, no. 8, pp. 30–37, Aug.2009.
    [3] S.Seo, J.Huang, H.Yang, andY.Liu, “Interpretable Convolutional Neural Networks with Dual Local and Global Attention for Review Rating Prediction,” in Proceedings of the Eleventh ACM Conference on Recommender Systems - RecSys ’17, 2017, pp. 297–305.
    [4] X.Wang, X.He, F.Feng, L.Nie, andT.-S.Chua, “TEM: Tree-enhanced Embedding Model for Explainable Recommendation,” Proc. 2018 World Wide Web Conf. World Wide Web - WWW ’18, no. 10, pp. 1543–1552, 2018.
    [5] J. L.Herlocker, J. A.Konstan, A.Borchers, andJ.Riedl, “An algorithmic framework for performing collaborative filtering,” in Proceedings of the 22nd annual international ACM SIGIR conference on Research and development in information retrieval - SIGIR ’99, 1999, pp. 230–237.
    [6] X. P.Yu, “Item-Based collaborative filtering recommendation algorithm,” Proc. 2009 Int. Conf. Mach. Learn. Cybern., vol. 3, pp. 1570–1576, 2009.
    [7] P. M. J. . B.D., “Hybrid Web Recommender Systems BT - The Adaptive Web: Methods and Strategies of Web Personalization,” 2007, pp. 377–408.
    [8] B.Hidasi, A.Karatzoglou, L.Baltrunas, andD.Tikk, “Session-based Recommendations with Recurrent Neural Networks,” pp. 1–10, 2015.
    [9] A.Paterek, “Improving regularized singular value decomposition for collaborative filtering,” Proc. KDD cup Work., pp. 2–5, 2007.
    [10] X.He, L.Liao, H.Zhang, L.Nie, X.Hu, andT.-S.Chua, “Neural Collaborative Filtering,” pp. 173–182, 2017.
    [11] Y.Hu, Y.Koren, andC.Volinsky, “Collaborative Filtering for Implicit Feedback Datasets,” in 2008 Eighth IEEE International Conference on Data Mining, 2008, pp. 263–272.
    [12] H.-T.Cheng et al., “Wide & Deep Learning for Recommender Systems,” pp. 1–4, 2016.
    [13] H. J.Xue, X. Y.Dai, J.Zhang, S.Huang, andJ.Chen, “Deep matrix factorization models for recommender systems,” IJCAI Int. Jt. Conf. Artif. Intell., pp. 3203–3209, 2017.
    [14] R.He andJ.McAuley, “VBPR: Visual Bayesian Personalized Ranking from Implicit Feedback,” pp. 144–150, 2015.
    [15] J.Chen, H.Zhang, X.He, L.Nie, W.Liu, andT.-S.Chua, “Attentive Collaborative Filtering: Multimedia Recommendation with Item- and Component-Level Attention,” Proc. 40th Int. ACM SIGIR Conf. Res. Dev. Inf. Retr. - SIGIR ’17, pp. 335–344, 2017.
    [16] D.Kim, C.Park, J.Oh, S.Lee, andH.Yu, “Convolutional Matrix Factorization for Document Context-Aware Recommendation,” pp. 233–240, 2016.
    [17] L.Baltrunas, B.Ludwig, andF.Ricci, “Matrix factorization techniques for context aware recommendation,” p. 301, 2011.
    [18] C.Bhagavatula, S.Feldman, R.Power, andW.Ammar, “Content-Based Citation Recommendation,” 2018.
    [19] S.Rendle, C.Freudenthaler, andL.Schmidt-Thieme, “Factorizing personalized Markov chains for next-basket recommendation,” p. 811, 2010.
    [20] B.Hidasi, M.Quadrana, A.Karatzoglou, andD.Tikk, “Parallel Recurrent Neural Network Architectures for Feature-rich Session-based Recommendations,” pp. 241–248, 2016.
    [21] J.Huang, W. X.Zhao, H.Dou, J.-R.Wen, andE. Y.Chang, “Improving Sequential Recommendation with Knowledge-Enhanced Memory Networks,” pp. 505–514, 2018.
    [22] J.Li, P.Ren, Z.Chen, Z.Ren, andJ.Ma, “Neural Attentive Session-based Recommendation,” pp. 1419–1428, 2017.
    [23] L.Zheng, V.Noroozi, andP. S.Yu, “Joint Deep Modeling of Users and Items Using Reviews for Recommendation,” pp. 425–433, 2017.
    [24] Q.Zhao et al., “Categorical-attributes-based item classification for recommender systems,” pp. 320–328, 2018.
    [25] G. C.-L.Hung, P.-C.Yang, C.-Y.Wang, andJ.-H.Chiang, “A Smartphone-Based Personalized Activity Recommender System for Patients with Depression,” in Proceedings of the 5th EAI International Conference on Wireless Mobile Communication and Healthcare - “Transforming healthcare through innovations in mobile and wireless technologies,” 2015, pp. 253–257.
    [26] Robin Burke, “Hybrid Recommender Systems: Survey and Experiments,” 2001.
    [27] H.Wang, N.Wang, andD.-Y.Yeung, “Collaborative Deep Learning for Recommender Systems,” 2014.
    [28] P.Vincent, H.Larochelle, I.Lajoie, Y.Bengio, andP.-A.Manzagol, Journal of machine learning research : JMLR., vol. 11. MIT Press, 2001.
    [29] O.Barkan andN.Koenigstein, “ITEM2VEC: Neural item embedding for collaborative filtering,” IEEE Int. Work. Mach. Learn. Signal Process. MLSP, vol. 2016-Novem, pp. 1–6, 2016.
    [30] T.Mikolov, I.Sutskever, K.Chen, G. S.Corrado, andJ.Dean, “Distributed Representations of Words and Phrases and their Compositionality.” pp. 3111–3119, 2013.
    [31] F.Murtagh, “Multilayer perceptrons for classification and regression,” Neurocomputing, vol. 2, no. 5–6, pp. 183–197, Jul.1991.
    [32] Q.Li, W.Cai, X.Wang, Y.Zhou, D. D.Feng, andM.Chen, “Medical image classification with convolutional neural network,” in 2014 13th International Conference on Control Automation Robotics & Vision (ICARCV), 2014, pp. 844–848.
    [33] Y.Kim, “Convolutional Neural Networks for Sentence Classification,” Aug.2014.
    [34] S. J.Pan andQ.Yang, “A Survey on Transfer Learning,” IEEE Trans. Knowl. Data Eng., vol. 22, no. 10, pp. 1345–1359, Oct.2010.
    [35] J. E.McDade, “Matrix Factorization Techniques for Recommender Systems,” J. Educ. Res., vol. 30, no. 7, pp. 489–503, 1937.
    [36] A.Van DenOord, S.Dieleman, andB.Schrauwen, “Deep content-based music recommendation.”
    [37] S.Rendle, C.Freudenthaler, Z.Gantner, andL.Schmidt-Thieme, “BPR: Bayesian Personalized Ranking from Implicit Feedback.”
    [38] Z.Li, H.Zhao, Q.Liu, Z.Huang, T.Mei, andE.Chen, “Learning from History and Present: Next-item Recommendation via Discriminatively Exploiting User Behaviors,” 2018.
    [39] X.Geng, H.Zhang, J.Bian, andT.-S.Chua, “Learning Image and User Features for Recommendation in Social Networks.”
    [40] D. M.Blei, A. Y.Ng, andM. I.Jordan, “Latent Dirichlet Allocation,” J. Mach. Learn. Res., vol. 3, no. Jan, pp. 993–1022, 2003.

    下載圖示 校內：2021-07-29公開
    校外：2021-07-29公開
    QR CODE