以支持向量機、優勢約略集合法與形式概念分析探勘生技美妝產品之消費者輪廓

Mining Consumer Behaviors of Biocosmetic Products by Using the Support Vector Machine, Dominance-based Rough Set Approach and Formal Concept Analysis

林奐均(Lin Huan-Chun)

指導教授 ： 黃啟祐

國立臺灣師範大學/科技與工程學院/工業教育學系/碩士(2022年)

https://doi.org/10.6345/NTNU202200721

摘要

隨著生技產業的發展，美妝產品也大量應用生技技術。隨著大數據時代的來臨，挖掘客戶行為模式已經成為日益盛行之行銷工具。精準的數據分析，能作為行銷人員訂定正確行銷策略，對於資源、經費有限的公司來說，不啻行銷利器。然而如何運用大數據分析，訂定生技美妝產品的行銷策略，少有學者研究。且如何解讀大數據分析所推衍之規則與結果，亦為生技美妝產業經理人與行銷人員亟待克服的問題。唯少有學者討論相關議題。
因此，本研究擬提出一新型決策分析框架，跨越研究缺口。本研究--使用支持向量機 (Support Vector Machine，SVM)、優勢約略集合法 (Dominance Based Rough Set Approach，DRSA)和形式概念分析(Formal Concept Analysis，FCA)探勘生技美妝產品之消費者行為，並採用人物誌 (Persona)，描繪出顧客的具體形象，幫助行銷人員清楚了解目標市場的客戶，設計出更符合其需求的產品。
為驗證分析架構之可行性，本研究導入國內某生技美妝公司於2018年至2019年間的客戶消費資料，並將消費者劃分為新客戶、沉睡客戶、流失客戶及 忠誠客戶等四個市場區隔後，導入優勢約略集合法，萃取每一市場區隔消費行為之規則，其後，利用形式概念分析法，歸納消費者行為，並將之視覺化。最後，使用人物誌描繪各市場區隔消費者之典型範例。本研究以Aiko、Hana、Rose與Sue等虛擬人物代表生技美妝市場四種市場區隔的客戶：Aiko喜歡購買與面部和嘴唇相關的化妝品和護膚品；Sue則喜歡購買爽膚水和乳液。經專家確認，四種虛擬人物與其消費行為，與實務經驗相符。本研究所提出的框架有助於識別客戶及其特徵，幫助行銷人員規劃策略。

關鍵字

生技美妝產品 ； 消費者行為 ； 大數據分析 ； 支持向量機 ； 優勢關係約略集合理論 ； 形式概念分析 ； 人物誌

並列摘要

With the development of the biotechnological industry, biotechnology is widely used in beauty products. With the advent of big data, mining customer behavior patterns has become an increasingly popular marketing tool. Marketing managers can use accurate data analysis to determine the correct marketing strategy. Primarily, data analytics can serve as a marketing tool for companies with limited resources and funds. However, few scholars have studied using big data analysis to determine marketing strategies for biotech beauty products. How to interpret the rules and results derived from big data analysis is also an urgent problem to be overcome by managers and marketers in the biotechnology beauty industry. Few scholars discuss the issue.
Therefore, this study proposes a new decision analysis framework to bridge the research gap. In this study, Support Vector Machine (SVM), Dominance Based Rough Set Approach, DRSA and Formal Concept Analysis (FCA) explore the consumer behavior of biotechnology beauty products, and use Persona to describe the specific image of customers, helping marketers clearly understand the target market customers, and design products that better meet their needs. SVM filters out the data with inconsistent characteristics. DRSA generates a generic description of each segmentation, the so-called decision rule. Those descriptions derive a concept hierarchy through example of segmenting consumers in each market is depicted with Personas.
In order to verify the feasibility of the analysis framework, this study introduces the customer consumption data of a domestic biotechnological beauty company from 2018 to 2019. The data is divided into four segments which are: new customers, sleep customers, lost customers, and loyal customers. The Virtual characters such as Aiko, Hana, Rose and Sue were used to represent customers in four market segments of biotech beauty market: Aiko likes to buy cosmetics and skincare products related to face and lips; Sue likes to buy toners and lotions. Confirmed by experts, four virtual characters and their consumption behavior, consistent with practical experience. The framework presented in this study helps identify customers and their characteristics and helps marketers plan their strategies.

並列關鍵字

Custommer behavior ； Dominance-based rough set approach ； Formal concept analysis ； Persona

參考文獻

Acharjya, D. P., & Ahmed, N. S. S. (2021). Tracing of online assaults in 5G networks using dominance based rough set and formal concept analysis. Peer-to-Peer Networking and Applications, 14(1), 349-374. doi:10.1007/s12083-020-00983-6

Ahmad, A. K., Jafar, A., & Aljoumaa, K. (2019). Customer churn prediction in telecom using machine learning in big data platform. Journal of Big Data, 6(1), 28. doi:10.1186/s40537-019-0191-6

Alves, A. L., Marques, A. L. P., Martins, E., Silva, T. H., & Reis, R. L. (2017). Cosmetic Potential of Marine Fish Skin Collagen. Cosmetics, 4(4). doi:10.3390/cosmetics4040039

An, J., Kwak, H., Jung, S.-g., Salminen, J., & Jansen, B. J. (2018). Customer segmentation using online platforms: isolating behavioral and demographic segments for persona creation via aggregated user data. Social Network Analysis and Mining, 8(1), 54. doi:10.1007/s13278-018-0531-0

Anitha, P., & Patil, M. M. (2019). RFM model for customer purchase behavior using K-Means algorithm. Journal of King Saud University - Computer and Information Sciences, 34(5), 1785-1792. doi:10.1016/j.jksuci.2019.12.011