臺灣博碩士論文加值系統

隨著現代科技與網際網路的發展，行動商務目前已逐漸成為人們生活方式的一環。而行動裝置的普及，也成為扮演著這場商務戰役當中不可或缺的一個重要角色。對於現今的服務應用中了解使用者的意圖將會是一個重要的議題，在本研究中，我們利用使用者手機查看、點擊及輸入的操作行為資料去推斷使用者的資訊調查意圖。基於此一觀點，於本論文中我們提出一架構，將調查意圖的問題轉換成序列資料二元分類問題。我們所提出之架構包含兩個主要元件：Episode擷取器(Episode Extractor)與意圖分類器(User Intention Classifier)。Episode擷取器主要探討如何於充滿雜訊的行動裝置使用行為中定義出最基本之序列單位，以此為輸入至後端之意圖分類器。而意圖分類器則探討利用遞迴類神經網路(RNN)演算法來建立行動使用者之意圖分類模型。實驗中，我們也透過實際之使用者資料收集來驗證我們所提出架構之效果，相較於最原始之方法，我們所提出調查意圖架構其準確率可達80%，證明我們架構之有效性。

Nowadays mobile devices have become a ubiquitous medium supporting various forms of functionality and are widely accepted for commons. With such an intimacy, a mobile device has been more than a mini computer for its owner but a personal behavior observer. As a role of behavior observer, mobile phones would be a new entry point to understand user's intention. Detecting a user's intention behind his/her activities is fundamental to many emerging commercial applications and intelligent services, such as recommendations making, targeted advertisements delivering, and personalized contents presenting. In this paper, we focus on capturing the state that a user is in his/her research phase. We refer to such a target as Information Research Intention (IRI). The intuition for capturing IRIs is that mobile phone usage activities containing IRIs might show sequential activity patterns, i.e., a type-in event is firstly observed, then a clicked event is followed, then a series of scrolling events is presented, and etc. By properly analyzing the mobile phone usage activities, detecting IRIs will be possible. Based on this viewpoint, in this thesis, we propose a framework that consists of two main components, Episode Extractor and User Intention Detection. The Episode extractor is designed to address the problem of extracting episodes from noisy, unstructured usage activity logs. And, the User Intention Detector is designed based on the idea of casting the detection problem as a binary classification problem. Experiments with real data collected from users is conducted to verify the effectiveness of the proposed framework for detecting IRI through context logs. From the experiment result, the accuracy of detection user intention is about 80%, which demonstrates the superiority of the proposed framework.

第1章緒論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 研究背景. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 研究動機與目的. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.3 主要貢獻. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.4 論文架構. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
第2章初步準備. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.1 行動裝置資料的蒐集. . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 資料模型與目標問題. . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 資料模型. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.2 目標問題. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Word Embedding Models . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.4 關鍵字擷取介紹. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
第3章研究方法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.1 方法架構及流程. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 Episode 擷取器. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.1 事件擷取. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.2 關鍵字擷取. . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2.3 事件向量表示與其之間的相似度. . . . . . . . . . . . . . . . 17
3.3 意圖分類器. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.3.1 定義向量序列資料. . . . . . . . . . . . . . . . . . . . . . . . 19
3.3.2 遞歸神經網路(Recurrent Neural Networks，RNN) . . . . . . . 19
3.3.3 分類器建構說明. . . . . . . . . . . . . . . . . . . . . . . . . . 22
第4章文獻探討. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.1 事件擷取. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.2 行為識別. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
4.3 引入序列分類模型於使用者意圖偵測. . . . . . . . . . . . . . . . . . 25
第5章實驗評估. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.1 實驗設定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.1.1 資料集. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.1.2 Episode 擷取器演算法評估. . . . . . . . . . . . . . . . . . . . 27
5.1.3 評估指標. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.2 Episode 擷取器之效能比較. . . . . . . . . . . . . . . . . . . . . . . . 28
5.3 意圖分類器之Episode 擷取器評估. . . . . . . . . . . . . . . . . . . . 29
5.4 意圖分類器之演算法評估. . . . . . . . . . . . . . . . . . . . . . . . . 32
5.5 Episode 擷取器參數評估. . . . . . . . . . . . . . . . . . . . . . . . . 32
5.5.1 基於時間Episode 擷取器參數評估. . . . . . . . . . . . . . . . 33
5.5.2 基於事件語義的Episode 擷取器之門檻值參數評估. . . . . . 33
5.5.3 DBKE 方法以及DBKE-CC 的參數評估. . . . . . . . . . . . . 34
5.5.4 TFIDF 方法的參數評估. . . . . . . . . . . . . . . . . . . . . . 36
5.6 實驗結果總結. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
第6章結論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
參考文獻. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

[1] T. Mikolov, K. Chen, G. Corrado, and J. Dean, “Efficient estimation of word repre- sentations in vector space,” arXiv preprint arXiv:1301.3781, 2013.
[2] Q. Le and T. Mikolov, “Distributed representations of sentences and documents,” in Proceedings of the 31st International Conference on Machine Learning (ICML-14), pp. 1188–1196, 2014.
[3] W. Ma, Y. Fang, and P. Lin, “Mobility management strategy based on user mobility patterns in wireless networks,” IEEE transactions on vehicular technology, vol. 56, no. 1, pp. 322–330, 2007.
[4] H.Abu-GhazalehandA.S.Alfa,“Applicationofmobilitypredictioninwirelessnet- works using markov renewal theory,” IEEE Transactions on Vehicular Technology, vol. 59, no. 2, pp. 788–802, 2010.
[5] O.Dousse,J.Eberle,andM.Mertens,“Placelearningviadirectwififingerprintclus- tering,” in Mobile Data Management (MDM), 2012 IEEE 13th International Con- ference on, pp. 282–287, IEEE, 2012.
[6] Y. Jiang, X. Pan, K. Li, Q. Lv, R. P. Dick, M. Hannigan, and L. Shang, “Ariel: Automatic wi-fi based room fingerprinting for indoor localization,” in Proceedings of the 2012 ACM Conference on Ubiquitous Computing, pp. 441–450, ACM, 2012.
[7] H.-H. Jo, M. Karsai, J. Kertész, and K. Kaski, “Circadian pattern and burstiness in mobile phone communication,” New Journal of Physics, vol. 14, no. 1, p. 013055, 2012.
[8] T.-P. Liang and C.-P. Wei, “Introduction to the special issue: Mobile commerce ap- plications,” International Journal of Electronic Commerce, vol. 8, no. 3, pp. 7–17, 2004.
[9] eMarketer, “Mobile Purchases to Account for a Third of UK Retail Ecommerce Sales.” https://www.emarketer.com/Article/ Mobile-Purchases-Account-Third-of-UK-Retail-Ecommerce-Sales/1012416, 2015. [Online; accessed 2017-07-18].
[10] H. K. Dai, L. Zhao, Z. Nie, J.-R. Wen, L. Wang, and Y. Li, “Detecting online com- mercial intention (oci),” in Proceedings of the 15th international conference on World Wide Web, pp. 829–837, ACM, 2006.
[11] S.HochreiterandJ.Schmidhuber,“Longshort-termmemory,”Neuralcomputation, vol. 9, no. 8, pp. 1735–1780, 1997.
[12] R. O. Duda, P. E. Hart, and D. G. Stork, Pattern classification. Wiley, New York, 1973.
[13] T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean, “Distributed repre- sentations of words and phrases and their compositionality,” in Advances in neural information processing systems, pp. 3111–3119, 2013.
[14] Apple, “Accessibility Programming Guide for iOS.” https://developer.apple.com/ library/content/documentation/UserExperience/Conceptual/iPhoneAccessibility/ Introduction/Introduction.html. [Online; accessed 2017-07-18].
[15] Google, “Android Accessibility API.” https://developer.android.com/guide/topics/ ui/accessibility/index.html. [Online; accessed 2017-07-18].
[16] G. Salton and C. Buckley, “Term-weighting approaches in automatic text retrieval,” Information processing management, vol. 24, no. 5, pp. 513–523, 1988.
[17] eMarketer, “ 巴 中 文 分 .” https://github.com/fxsjy/jieba, 2014. [Online; ac- cessed 2017-07-18].
[18] M. Ester, H.-P. Kriegel, J. Sander, X. Xu, et al., “A density-based algorithm for discovering clusters in large spatial databases with noise.,” in Kdd, vol. 96, pp. 226– 231, 1996.
[19] A. Graves, “Generating sequences with recurrent neural networks,” arXiv preprint arXiv:1308.0850, 2013.
[20] I. Sutskever, O. Vinyals, and Q. V. Le, “Sequence to sequence learning with neural networks,” in Advances in neural information processing systems, pp. 3104–3112, 2014.
[21] I. Sutskever, J. Martens, and G. E. Hinton, “Generating text with recurrent neural networks,” in Proceedings of the 28th International Conference on Machine Learning (ICML-11), pp. 1017–1024, 2011.
[22] J. L. Elman, “Finding structure in time,” Cognitive science, vol. 14, no. 2, pp. 179– 211, 1990.
[23] L. D. Catledge and J. E. Pitkow, “Characterizing browsing strategies in the world- wide web,” Computer Networks and ISDN systems, vol. 27, no. 6, pp. 1065–1073, 1995.
[24] H. Cao, D. Jiang, J. Pei, Q. He, Z. Liao, E. Chen, and H. Li, “Context-aware query suggestion by mining click-through and session data,” in Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining, pp. 875–883, ACM, 2008.
[25] Z. Liao, Y. Song, L.-w. He, and Y. Huang, “Evaluating the effectiveness of search task trails,” in Proceedings of the 21st international conference on World Wide Web, pp. 489–498, ACM, 2012.
[26] R. W. White, P. N. Bennett, and S. T. Dumais, “Predicting short-term interests using activity-based search context,” in Proceedings of the 19th ACM international con- ference on Information and knowledge management, pp. 1009–1018, ACM, 2010.
[27] B. Xiang, D. Jiang, J. Pei, X. Sun, E. Chen, and H. Li, “Context-aware ranking in web search,” in Proceedings of the 33rd international ACM SIGIR conference on Research and development in information retrieval, pp. 451–458, ACM, 2010.
[28] D.HeandA.Göker,“Detectingsessionboundariesfromwebuserlogs,”inProceed- ings of the BCS-IRSG 22nd annual colloquium on information retrieval research, pp. 57–66, 2000.
[29] R. Jones and K. L. Klinkner, “Beyond the session timeout: automatic hierarchical segmentation of search topics in query logs,” in Proceedings of the 17th ACM con- ference on Information and knowledge management, pp. 699–708, ACM, 2008.
[30] K. Bollacker, C. Evans, P. Paritosh, T. Sturge, and J. Taylor, “Freebase: a col- laboratively created graph database for structuring human knowledge,” in Proceed- ings of the 2008 ACM SIGMOD international conference on Management of data, pp. 1247–1250, AcM, 2008.
[31] J. Cui, H. Liu, J. Yan, L. Ji, R. Jin, J. He, Y. Gu, Z. Chen, and X. Du, “Multi-view random walk framework for search task discovery from click-through log,” in Pro- ceedings of the 20th ACM international conference on Information and knowledge management, pp. 135–140, ACM, 2011.
[32] J. Huang and E. N. Efthimiadis, “Analyzing and evaluating query reformulation strategies in web search logs,” in Proceedings of the 18th ACM conference on In- formation and knowledge management, pp. 77–86, ACM, 2009.
[33] C. Lucchese, S. Orlando, R. Perego, F. Silvestri, and G. Tolomei, “Identifying task- based sessions in search engine query logs,” in Proceedings of the fourth ACM in- ternational conference on Web search and data mining, pp. 277–286, ACM, 2011.
[34] M. Kamvar and S. Baluja, “A large scale study of wireless search behavior: Google mobile search,” in Proceedings of the SIGCHI conference on Human Factors in com- puting systems, pp. 701–709, ACM, 2006.
[35] K.Church,B.Smyth,P.Cotter,andK.Bradley,“Mobileinformationaccess:Astudy of emerging search behavior on the mobile internet,” ACM Transactions on the Web (TWEB), vol. 1, no. 1, p. 4, 2007.
[36] M.KamvarandS.Baluja,“Decipheringtrendsinmobilesearch,”Computer,vol.40, no. 8, 2007.
[37] L. R. Rabiner, “A tutorial on hidden markov models and selected applications in speech recognition,” Proceedings of the IEEE, vol. 77, no. 2, pp. 257–286, 1989.
[38] P. Pook and D. Ballard, “Recognizing teleoperating manipulations,” in International Conference on Robotics and Automation, pp. 578–585.
[39] G. E. Hovland, P. Sikka, and B. J. McCarragher, “Skill acquisition from human demonstration using a hidden markov model,” in Robotics and Automation, 1996. Proceedings., 1996 IEEE International Conference on, vol. 3, pp. 2706–2711, Ieee, 1996.
[40] K. Ogawara, J. Takamatsu, H. Kimura, and K. Ikeuchi, “Modeling manipulation interactions by hidden markov models,” in Intelligent Robots and Systems, 2002. IEEE/RSJ International Conference on, vol. 2, pp. 1096–1101, IEEE, 2002.
[41] R. Kelley, A. Tavakkoli, C. King, A. Ambardekar, M. Nicolescu, and M. Nicolescu, “Context-based bayesian intent recognition,” IEEE Transactions on Autonomous Mental Development, vol. 4, no. 3, pp. 215–225, 2012.