研究動機與目的：精準行銷成效取決於消費者標籤的精準程度，而消費者分群效果受到序列資料特徵、所使用的距離計算方法等因素影響；本研究以實務案例驗證，導入不同的序列距離計算方式，是否有助於萃取消費者行為資訊差異並優化其分群效果。

研究方法：透過狀態序列轉換後，導入最佳匹配法計算序列相異度矩陣，最後分群觀察序列狀態分布比例圖表，並計算兩群間特徵指標統計顯著性，以驗證分群所得之群體特徵具有顯著差異。

研究應用場景：只要有時間戳記的日誌格式資料，透過狀態定義形成序列後，皆可以導入最佳匹配法運算出特徵，並用於分群、分類等不同算法中。

研究貢獻：本研究以「電商用戶行為序列分群」及「物流進程序列分群」兩案例，實證最佳匹配法所運算產生之特徵，相較於事件次數累計，更有助於提高分群多樣性，且能使分群群體特徵指標間產生顯著差異。

Motivation and purpose: Ecommerce advertising retargeting performance highly depends on the quality of audience cluster tagging, and the cluster quality is affected by different sequence features transform methods. Our research aimed to validate that whether using optimal matching to generate sequence dissimilarity increased the diversity and significance of clustering results. Furthermore, we also discussed sequence clustering use cases in delivery logistics satisfaction to compare different scenario of sequence clustering.

Research method: First, we convert log data to state sequences, then compute sequence dissimilarity matrix using optimal matching algorithm. Last, run clustering algorithm and observe the state distribution plot among different clusters, with the results of Kruskal-Wallis significance test to validate that significant difference exists between key metrics of those two clusters.

Implement scenario: As long as there’s log format data with timestamps, we can transform it into state sequences through state definition, then generate dissimilarity matrix as a feature used in clustering, classification and other algorithms for increasing performance.

Research value: Our research validated that generating sequence dissimilarity by optimal matching algorithm not only increased the diversity of clustering results with more user pattern observed, but also segmented significantly different types of clusters using only state sequences data. Besides, we perform two analyses to show the entire process from data transformation, modeling and visualization.

第一章　緒論　7
第二章　文獻探討　10
第一節　電子商務用戶行為序列格式與分群應用　10
第二節　序列距離計算方式　13
第三節　分群演算法　15
第三章　最佳匹配法用於序列分群　17
第一節　OM 序列相異量估計：置換成本、增刪成本　19
第二節　OM 演算法之變化型　20
第三節　狀態序列分析研究流程　26
第四章　資料分析案例一：電子商務用戶行為序列分群　30
第一節　資料前處理與狀態序列轉換　30
第二節　狀態序列分群視覺化與效果衡量　36
第五章　資料分析案例二：訂單物流進程序列分群　51
第一節　資料前處理與狀態序列轉換　51
第二節　狀態序列分群視覺化與效果衡量　56
第六章　研究價值與未來展望　62
參考文獻　65

[1] 普華永道會計師事務所。(民 108 年)。2019 全球消費者洞察報告。民 109 年 6 月 28 日，取自:https://www.pwc.tw/zh/publications/global-insights/2019-consumer- insights.html
[2] Google/Ipsos. (2017). Shopping Tracker. Retrieved June 28, 2020 from: https://www.thinkwithgoogle.com/data/us-shopping-behavior-statistics/
[3] Chen, Y., Fan, C., Li, Z., & Ren, L. (2020). Research on the Relationship between Precision Marketing and Company Development Ability. 2nd International Conference on Big-data Service and Intelligent Computation (BDSIC 2020). Association for Computing Machinery, New York, NY, USA, 42–48. DOI: https://doi.org/10.1145/3440054.3440062
[4] 劉洪偉、梁周揚、左妹華、陸丹、范夢婷、何銳超(民 108 年)。利用消費者瀏 覽行為識別品牌競爭關係研究。廣東工業大學學報，36(05)，1-6，13。DOI:
https://doi.org/10.12052/gdutxb.190063
[5] Verheijden, R.M.C. (2012). Predicting purchasing behavior throughout the clickstream. DOI: https://research.tue.nl/en/studentTheses/predicting-purchasing-behavior-throughout-the- clickstream
[6] 何銳超、劉洪偉、高鴻銘、范夢婷、詹明君(民 109 年)。基於點擊流與 PROMETHEE多屬性決策法的電子商務消費者購買意願預測。廣東工業大學學報， 37(06)，32-40。DOI: http://dx.doi.org/10.12052/gdutxb.200029
[7] Koehn, D., Lessmann, S., & Schaal, M. (2020). Predicting online shopping behaviour from clickstream data using deep learning. Expert Systems with Applications, Vol. 150, 2020, 113342, ISSN 0957-4174. DOI: https://doi.org/10.1016/j.eswa.2020.113342
[8] Wang, G., Zhang, X., Tang, S., Zheng, H., & Zhao, B. Y. (2016). Unsupervised Clickstream Clustering for User Behavior Analysis. In Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems (CHI '16). Association for Computing Machinery, New York, NY, USA, 225–236. DOI: https://doi.org/10.1145/2858036.2858107
[9] 張鴻萍(民 106 年)。基於時間序列交易數據的服裝電商客戶分類研究。現代管理，7，6。DOI: https://m.hanspub.org/journal/paper/23238#ref2
[10] Su, Q., & Chen, L. (2015). A method for discovering clusters of e-commerce interest patterns using click-stream data. Electronic Commerce Research and Applications, Vol. 14, Issue 1, 2015, 1-13, ISSN 1567-4223. DOI: https://doi.org/10.1016/j.elerap.2014.10.002
[11] Wei, J., Shen, Z., Sundaresan, N., & Ma, K. L. (2012). Visual cluster exploration of web clickstream data. 2012 IEEE Conference on Visual Analytics Science and Technology (VAST), Seattle, WA, USA, 2012, 3-12, DOI: https://doi.org/10.1109/VAST.2012.6400494
[12] Luu, V. T., Forestier, G., Weber, J., Bourgeois, P., Djelil, F & Muller, P. A. (2020). A review of alignment based similarity measures for web usage mining. Artif Intell
Rev 53, 1529–1551. DOI: https://doi.org/10.1007/s10462-019-09712-9
[13] Mandal, O.P., & Azad, H. K. (2014). Web Access Prediction Model using Clustering and Artificial Neural Network. International journal of engineering research and technology, 3. DOI: https://api.semanticscholar.org/CorpusID:61917769.
[14] Deza, M. M., & Deza, E. (2013). Distances and Similarities in Data Analysis. In: Encyclopedia of Distances. Springer, Berlin, Heidelberg. DOI: https://doi.org/10.1007/978-3- 642-30958-8_17
[15] Poornalatha, G., & Raghavendra, P. S. (2011). Web User Session Clustering Using Modified K-Means Algorithm. In: Abraham A., Lloret Mauri J., Buford J.F., Suzuki J., Thampi S.M. (eds) Advances in Computing and Communications. ACC 2011. Communications in Computer and Information Science, Vol. 191. Springer, Berlin, Heidelberg. DOI: https://doi.org/10.1007/978-3-642-22714-1_26
[16] Needleman, S. B., & Wunsch, C. D. (1970). A general method applicable to the search for similarities in the amino acid sequence of two proteins. Journal of molecular biology, 48(03), 443–453. DOI: https://doi.org/10.1016/0022-2836(70)90057-4
[17] Abbott, A., & Forrest, J. (1986). Optimal Matching Methods for Historical Sequences. The Journal of Interdisciplinary History, 16(3), 471-494. DOI: https://doi.org/10.2307/204500
[18] Flöthmann, C., & Hoberg, K. (2017). Career Patterns of Supply Chain Executives: An Optimal Matching Analysis. Journal of Business Logistics, 38, 35-54. DOI: https://doi.org/10.1111/jbl.12150
[19] Kaushik, M., & Mathur, B. (2014). Comparative Study of K-Means and Hierarchical Clustering Techniques. Internatioonal journal of Software and Hardware Research in Engineering, 2, 6, 93-98. DOI: http://ijournals.in/wp-content/uploads/2017/07/IJSHRE- 2653.compressed.pdf
[20] Roger, J(無日期)。3-2 Hierarchical Clustering(階層式分群法)。Data Clustering and Pattern Recognition(資料分群與樣式辨認)。Chap. 3-2。民 110 年 2 月 7 日，取自:http://mirlab.org/jang/books/dcpr/dcHierClustering.asp?title=3- 2%20Hierarchical%20Clustering%20(%B6%A5%BCh%A6%A1%A4%C0%B8s%AAk)&l anguage=chinese
[21] 卓明宏(無日期)。機率式重分配的模擬退火 K-means 演算法。民 110 年 2 月 7 日，取自:https://ir.nctu.edu.tw/bitstream/11536/78190/2/355203.pdf
[22] Govender, P., & Sivakumar, V. (2020). Application of k-means and hierarchical clustering techniques for analysis of air pollution: A review (1980–2019). Atmospheric Pollution Research, Vol 11, Issue 1, 40–56, ISSN 1309-1042, DOI: https://doi.org/10.1016/j.apr.2019.09.009
[23] Ward, J. H. (1963). Hierarchical Grouping to Optimize an Objective Function. Journal of the American Statistical Association, 58:301, 236–
244, DOI: https://doi.org/10.1080/01621459.1963.10500845
[24] K-平均算法(無日期)。維基百科。民 110 年 2 月 7 日，取自: https://zh.wikipedia.org/w/index.php?title=K- %E5%B9%B3%E5%9D%87%E7%AE%97%E6%B3%95&oldid=64158237
[25] Loss Dragon(民 108 年 1 月 18 日)。數據挖掘入門筆記—K-Medoids【投 稿】。知乎專欄。民 110 年 3 月 20 日，取自:https://zhuanlan.zhihu.com/p/55163617
[26] Studer, M., & Ritschard, G. (2016), What matters in differences between life trajectories: a comparative review of sequence dissimilarity measures. J. R. Stat. Soc. A, 179: 481-511. DOI: https://doi.org/10.1111/rssa.12125
[27] Gabadinho, A., Ritschard, G., Müller, N., & Studer, M. (2011). Analyzing and Visualizing State Sequences in R with TraMineR. Journal of Statistical Software, 40(4), 1 - 37. DOI: http://dx.doi.org/10.18637/jss.v040.i04
[28] Hollister, M. (2009). Is Optimal Matching Suboptimal? Sociological Methods & Research, 38(2), 235–264. DOI: https://doi.org/10.1177/0049124109346164
[29] Halpin, B. (2010). Optimal Matching Analysis and Life-Course Data: The Importance of Duration. Sociological Methods & Research, 38(3), 365–388. DOI: https://doi.org/10.1177/0049124110363590
[30] Biemann, T. (2011). A Transition-Oriented Approach to Optimal Matching. Sociological Methodology, 41(1), 195–221. DOI: https://doi.org/10.1111/j.1467- 9531.2011.01235.x
[31] Rousseeuw, P. J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics, Vol. 20, 53-65, ISSN 0377-0427, DOI: https://doi.org/10.1016/0377-0427(87)90125-7
[32] 天貓複購預測之挑戰 Baseline – 阿里雲天池(每賽季更新)【資料檔】。天貓商 城 Tmall.com。民 108 年 6 月 1 日，取自: https://tianchi.aliyun.com/competition/entrance/231576/information
[33] Dlouhy, K., & Biemann, T. (2015). Optimal Matching Analysis in Career Research: A Review and Some Best-practice Recommendations. Journal of Vocational Behavior, 90, 163- 173. DOI: https://doi.org/10.1016/J.JVB.2015.04.005