學術產出-學位論文
文章檢視/開啟
書目匯出
-
題名 社群協同合作平台之推薦問題研究－以GitHub為例
A study of recommendations on social collaboration platforms : using GitHub as an example作者 崔嘉祐
Tsui, Chia-Yu貢獻者 蔡銘峰<br>王釧茹
Tsai, Ming-Feng<br>Wang, Chuan-Ju
崔嘉祐
Tsui, Chia-Yu關鍵詞 推薦
協同過濾
基於內容過濾
分解機器
Recommendation
Collaborative filtering
Content-based filtering
Factorizaction machine日期 2017 上傳時間 2-十月-2017 10:16:50 (UTC+8) 摘要 本論文提出一種在社群協作平台 GitHub 上的推薦方法,利用社群 協作平台上的資訊於分解機器 ( Factorizaction Machines ,簡稱 FM ) 模 型中。 首先,我們抽取專案的協作關係、專案內的文字與程式碼,當 作是特徵資訊加入模型中訓練,進而以模型的訓練結果去做推薦。我 們利用 GitHub 平台上, 開發者對專案的行為 ( 如,給予的星號、關 注、開分支編輯與貢獻 ) ,去建立開發者對專案感興趣評級,並產生 使用者與專案的關係矩陣來當作我們的學習目標。 藉由這樣的方法, 我們不僅能夠幫助模型收斂還能提升推薦的結果,而這種透過不同 數量的相似特徵方法還可幫助使用者接觸到更多面向的物品。在實驗 中, 不論是加入文字特徵還是程式碼特徵, 相較於傳統的推薦方法協 同過濾,我們在平均精確均值 ( Mean Average Precision, MAP) 、 召回 率 ( Recall ) 與 F1 分數 ( F1 score ) 三個評估下都有較優秀的表現。 最 後,實驗結果顯示,在這種協作開發專案的 GitHub 社群協作平台上, 除了一般文字資訊外,程式碼資訊在推薦上是更有幫助的特徵資訊。
This paper proposes a recommendation approach based on Factorization Machines (FM) for GitHub, a social collaborative platform for program de- velopment. This work first extracts several features related to collaboration relationship and textual information within the project and the codes, and then incorporates the features into the model for training and learning. Lastly, the learned models are utilized for recommendation. This work skillfully uti- lizes the behaviors of developers toward a project, such as the star labeling, watch, fork, and contribution, to establish the degree of interest of a devel- oper has toward a project. Then, the proposed approach follows the con- struction of User-Item matrix for conducting the FM learning process. This approach not only expedites the convergence speed and the accuracy of FM, but it also enables users to explore the objects from different aspects. In the experiments, we compare the proposed approach with the traditional collab- oration filtering methods in terms of Mean Average Precision (MAP), Recall and F1 measures. The experimental results show that the proposed method outperforms the traditional user-based and item-based collaboration filtering methods. Furthermore, the experiment shows that, for social collaboration platform for program development, the incorporation of code feature is of greater enhancement than textual feature in the task of recommendation.參考文獻 [1] F. Aiolli. A preliminary study on a recommender system for the million songs dataset challenge. In Proceedings of the 2013 International Institute of Refriger- ation (IIR), pages 73–83. Citeseer, 2013.[2] P. Bedi, H. Kaur, and S. Marwaha. Trust based recommender system for seman- tic web. In Proceedings of the 2007 International Joint Conference on Artificial Intelligence (IJCAI), volume 7, pages 2677–2682, 2007.[3] M. Cha, A. Mislove, and K. P. Gummadi. A measurement-driven analysis of infor- mation propagation in the flickr social network. In Proceedings of the 18th Interna- tional Conference on World Wide Web, pages 721–730. ACM, 2009.[4] J. Davidson, B. Liebald, J. Liu, P. Nandy, T. Van Vleet, U. Gargi, S. Gupta, Y. He, M. Lambert, B. Livingston, et al. The youtube video recommendation system. In Proceedings of the 4th ACM Conference on Recommender Systems, pages 293–296. ACM, 2010.[5] D. Goldberg, D. Nichols, B. M. Oki, and D. Terry. Using collaborative filtering to weave an information tapestry. Communications of the ACM, 35(12):61–70, 1992.[6] D. Gruhl, R. Guha, D. Liben-Nowell, and A. Tomkins. Information diffusion through blogspace. In Proceedings of the 13th International Conference on World Wide Web, pages 491–501. ACM, 2004.[7] L. Hong, O. Dan, and B. D. Davison. Predicting popular messages in twitter. In Proceedings of the 20th International Conference Companion on World Wide Web, pages 57–58. ACM, 2011.[8] L. Hong, A. S. Doumith, and B. D. Davison. Co-factorization machines: modeling user interests and predicting individual decisions in twitter. In Proceedings of the 6th ACM International Conference on Web Search and Data Mining, pages 557– 566. ACM, 2013.[9] M. Jiang, P. Cui, R. Liu, Q. Yang, F. Wang, W. Zhu, and S. Yang. Social contex- tual recommendation. In Proceedings of the 21st ACM International Conference on Information and Knowledge Management, pages 45–54. ACM, 2012.[10] G. Linden, B. Smith, and J. York. Amazon. com recommendations: Item-to-item collaborative filtering. IEEE Internet Computing, 7(1):76–80, 2003.[11] H. Ma, I. King, and M. R. Lyu. Learning to recommend with social trust ensemble. In Proceedings of the 32nd International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 203–210. ACM, 2009.[12] P. Massa and P. Avesani. Trust-aware recommender systems. In Proceedings of the 2007 ACM Conference on Recommender Systems, pages 17–24. ACM, 2007.[13] I. Pila ́szy and D. Tikk. Recommending new movies: even a few ratings are more valuable than metadata. In Proceedings of the 3rd ACM Conference on Recom- mender Systems, pages 93–100. ACM, 2009.[14] S. Rendle. Factorization machines. In Processing of the 10th IEEE International Conference on Data Mining (ICDM), pages 995–1000. IEEE, 2010.[15] S. Rendle. Factorization machines with libfm. Proceedings of the 2012 ACM Trans- actions on Intelligent Systems and Technology (TIST), 3(3):57, 2012.[16] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl. Item-based collaborative filtering recommendation algorithms. In Proceedings of the 10th International Conference on World Wide Web, pages 285–295. ACM, 2001.[17] J. Schafer, D. Frankowski, J. Herlocker, and S. Sen. Collaborative filtering recom- mender systems. The adaptive web, pages 291–324, 2007.[18] Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson, A. Hanjalic, and N. Oliver. Tfmap: optimizing map for top-n context-aware recommendation. In Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Infor- mation Retrieval, pages 155–164. ACM, 2012.[19] J. Weston, C. Wang, R. Weiss, and A. Berenzweig. Latent collaborative retrieval. arXiv preprint arXiv:1206.4603, 2012. 描述 碩士
國立政治大學
資訊科學學系
104753034資料來源 http://thesis.lib.nccu.edu.tw/record/#G0104753034 資料類型 thesis dc.contributor.advisor 蔡銘峰<br>王釧茹 zh_TW dc.contributor.advisor Tsai, Ming-Feng<br>Wang, Chuan-Ju en_US dc.contributor.author (作者) 崔嘉祐 zh_TW dc.contributor.author (作者) Tsui, Chia-Yu en_US dc.creator (作者) 崔嘉祐 zh_TW dc.creator (作者) Tsui, Chia-Yu en_US dc.date (日期) 2017 en_US dc.date.accessioned 2-十月-2017 10:16:50 (UTC+8) - dc.date.available 2-十月-2017 10:16:50 (UTC+8) - dc.date.issued (上傳時間) 2-十月-2017 10:16:50 (UTC+8) - dc.identifier (其他 識別碼) G0104753034 en_US dc.identifier.uri (URI) http://nccur.lib.nccu.edu.tw/handle/140.119/113296 - dc.description (描述) 碩士 zh_TW dc.description (描述) 國立政治大學 zh_TW dc.description (描述) 資訊科學學系 zh_TW dc.description (描述) 104753034 zh_TW dc.description.abstract (摘要) 本論文提出一種在社群協作平台 GitHub 上的推薦方法,利用社群 協作平台上的資訊於分解機器 ( Factorizaction Machines ,簡稱 FM ) 模 型中。 首先,我們抽取專案的協作關係、專案內的文字與程式碼,當 作是特徵資訊加入模型中訓練,進而以模型的訓練結果去做推薦。我 們利用 GitHub 平台上, 開發者對專案的行為 ( 如,給予的星號、關 注、開分支編輯與貢獻 ) ,去建立開發者對專案感興趣評級,並產生 使用者與專案的關係矩陣來當作我們的學習目標。 藉由這樣的方法, 我們不僅能夠幫助模型收斂還能提升推薦的結果,而這種透過不同 數量的相似特徵方法還可幫助使用者接觸到更多面向的物品。在實驗 中, 不論是加入文字特徵還是程式碼特徵, 相較於傳統的推薦方法協 同過濾,我們在平均精確均值 ( Mean Average Precision, MAP) 、 召回 率 ( Recall ) 與 F1 分數 ( F1 score ) 三個評估下都有較優秀的表現。 最 後,實驗結果顯示,在這種協作開發專案的 GitHub 社群協作平台上, 除了一般文字資訊外,程式碼資訊在推薦上是更有幫助的特徵資訊。 zh_TW dc.description.abstract (摘要) This paper proposes a recommendation approach based on Factorization Machines (FM) for GitHub, a social collaborative platform for program de- velopment. This work first extracts several features related to collaboration relationship and textual information within the project and the codes, and then incorporates the features into the model for training and learning. Lastly, the learned models are utilized for recommendation. This work skillfully uti- lizes the behaviors of developers toward a project, such as the star labeling, watch, fork, and contribution, to establish the degree of interest of a devel- oper has toward a project. Then, the proposed approach follows the con- struction of User-Item matrix for conducting the FM learning process. This approach not only expedites the convergence speed and the accuracy of FM, but it also enables users to explore the objects from different aspects. In the experiments, we compare the proposed approach with the traditional collab- oration filtering methods in terms of Mean Average Precision (MAP), Recall and F1 measures. The experimental results show that the proposed method outperforms the traditional user-based and item-based collaboration filtering methods. Furthermore, the experiment shows that, for social collaboration platform for program development, the incorporation of code feature is of greater enhancement than textual feature in the task of recommendation. en_US dc.description.tableofcontents 第一章介紹 11.1 前言 11.2 研究目的 2第二章相關文獻探討 32.1 推薦系統( Recommendation Systems ) 32.1.1 協同過濾( Collaborative Filtering ) 32.1.2 基於內容過濾( Content-Based Filtering ) 42.1.3 混合型演算法 42.2 分解機器( Factorization Machines ) 5第三章研究方法 73.1 一般的分解機器( Standard Factorization Machine ) 73.2 社群協作平台的推薦框架 73.3 基於內容特徵( Content-based Feature ) 93.3.1 一般文字的特徵資訊 103.3.2 程式碼的特徵資訊 11第四章實驗結果與討論 134.1 實驗設定 134.1.1 資料蒐集 134.1.2 實驗資料集 144.1.3 評估指標 154.2 社群協作平台推薦系統 154.2.1 基於協同過濾的推薦 164.2.2 分解機器加入文字特徵 164.2.3 分解機器加入程式碼特徵 174.2.4 分解機器加入文字特徵與程式碼特徵 174.3 敏感度測試 184.3.1 隱向量的維度k 184.3.2 模型執行迴圈次數 214.4 實驗結果之改善重要性檢測 224.5 特徵權重 22第五章結論 24 zh_TW dc.format.extent 1921187 bytes - dc.format.mimetype application/pdf - dc.source.uri (資料來源) http://thesis.lib.nccu.edu.tw/record/#G0104753034 en_US dc.subject (關鍵詞) 推薦 zh_TW dc.subject (關鍵詞) 協同過濾 zh_TW dc.subject (關鍵詞) 基於內容過濾 zh_TW dc.subject (關鍵詞) 分解機器 zh_TW dc.subject (關鍵詞) Recommendation en_US dc.subject (關鍵詞) Collaborative filtering en_US dc.subject (關鍵詞) Content-based filtering en_US dc.subject (關鍵詞) Factorizaction machine en_US dc.title (題名) 社群協同合作平台之推薦問題研究－以GitHub為例 zh_TW dc.title (題名) A study of recommendations on social collaboration platforms : using GitHub as an example en_US dc.type (資料類型) thesis en_US dc.relation.reference (參考文獻) [1] F. Aiolli. A preliminary study on a recommender system for the million songs dataset challenge. In Proceedings of the 2013 International Institute of Refriger- ation (IIR), pages 73–83. Citeseer, 2013.[2] P. Bedi, H. Kaur, and S. Marwaha. Trust based recommender system for seman- tic web. In Proceedings of the 2007 International Joint Conference on Artificial Intelligence (IJCAI), volume 7, pages 2677–2682, 2007.[3] M. Cha, A. Mislove, and K. P. Gummadi. A measurement-driven analysis of infor- mation propagation in the flickr social network. In Proceedings of the 18th Interna- tional Conference on World Wide Web, pages 721–730. ACM, 2009.[4] J. Davidson, B. Liebald, J. Liu, P. Nandy, T. Van Vleet, U. Gargi, S. Gupta, Y. He, M. Lambert, B. Livingston, et al. The youtube video recommendation system. In Proceedings of the 4th ACM Conference on Recommender Systems, pages 293–296. ACM, 2010.[5] D. Goldberg, D. Nichols, B. M. Oki, and D. Terry. Using collaborative filtering to weave an information tapestry. Communications of the ACM, 35(12):61–70, 1992.[6] D. Gruhl, R. Guha, D. Liben-Nowell, and A. Tomkins. Information diffusion through blogspace. In Proceedings of the 13th International Conference on World Wide Web, pages 491–501. ACM, 2004.[7] L. Hong, O. Dan, and B. D. Davison. Predicting popular messages in twitter. In Proceedings of the 20th International Conference Companion on World Wide Web, pages 57–58. ACM, 2011.[8] L. Hong, A. S. Doumith, and B. D. Davison. Co-factorization machines: modeling user interests and predicting individual decisions in twitter. In Proceedings of the 6th ACM International Conference on Web Search and Data Mining, pages 557– 566. ACM, 2013.[9] M. Jiang, P. Cui, R. Liu, Q. Yang, F. Wang, W. Zhu, and S. Yang. Social contex- tual recommendation. In Proceedings of the 21st ACM International Conference on Information and Knowledge Management, pages 45–54. ACM, 2012.[10] G. Linden, B. Smith, and J. York. Amazon. com recommendations: Item-to-item collaborative filtering. IEEE Internet Computing, 7(1):76–80, 2003.[11] H. Ma, I. King, and M. R. Lyu. Learning to recommend with social trust ensemble. In Proceedings of the 32nd International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 203–210. ACM, 2009.[12] P. Massa and P. Avesani. Trust-aware recommender systems. In Proceedings of the 2007 ACM Conference on Recommender Systems, pages 17–24. ACM, 2007.[13] I. Pila ́szy and D. Tikk. Recommending new movies: even a few ratings are more valuable than metadata. In Proceedings of the 3rd ACM Conference on Recom- mender Systems, pages 93–100. ACM, 2009.[14] S. Rendle. Factorization machines. In Processing of the 10th IEEE International Conference on Data Mining (ICDM), pages 995–1000. IEEE, 2010.[15] S. Rendle. Factorization machines with libfm. Proceedings of the 2012 ACM Trans- actions on Intelligent Systems and Technology (TIST), 3(3):57, 2012.[16] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl. Item-based collaborative filtering recommendation algorithms. In Proceedings of the 10th International Conference on World Wide Web, pages 285–295. ACM, 2001.[17] J. Schafer, D. Frankowski, J. Herlocker, and S. Sen. Collaborative filtering recom- mender systems. The adaptive web, pages 291–324, 2007.[18] Y. Shi, A. Karatzoglou, L. Baltrunas, M. Larson, A. Hanjalic, and N. Oliver. Tfmap: optimizing map for top-n context-aware recommendation. In Proceedings of the 35th International ACM SIGIR Conference on Research and Development in Infor- mation Retrieval, pages 155–164. ACM, 2012.[19] J. Weston, C. Wang, R. Weiss, and A. Berenzweig. Latent collaborative retrieval. arXiv preprint arXiv:1206.4603, 2012. zh_TW
