結合科技接受模式與資料採礦方法進行智慧型電視之購買預測

Combining Technology Acceptance Model with Data Mining Methods in Smart-TV Users’ Purchase Forecasting

陳澤明(Ze-Ming Chen)

指導教授 ： 王志軒

國立交通大學/管理學院/工業工程與管理系所/碩士(2015年)

https://doi.org/10.6842/NCTU.2015.00455

摘要

電視原本就是人類生活中重要的裝置，而近年來隨著智慧型產品的風潮，智慧型電視也被推出到市場。且智慧型電視因為有著大螢幕、高運算儲存能力等優勢，使其將來在物聯網、智慧家庭的概念中，均有擔任控制中樞的潛力。目前智慧型電視市場還在起步階段，市面上的產品皆是第一、二代的產品，功能與介面的歧異度均大。由2015的Hype cycle週期可以知道智慧型電視目前處在期望下降區，並對照新產品擴散的貝氏理論，不難發現現在智慧型電視的消費者主要創新者。而在創新者以後，才要迎入占了產品消費者九成以上的模仿者。這些模仿者對於3C產品的熱誠與好奇心較創新者少了許多，因此如何搶在其他製造商前推出能夠滿足創新者且能吸引模仿者的主流產品成了目前智慧型電視製造商最重要的課題。
本研究先以羅吉斯迴歸、粗糙集以及支撐向量機進行功能手勢變數的購買預測，在確定了模型的有效性及一致性以後，更進一步的以粗糙集以及羅吉斯迴歸搜尋關鍵的變數。接著為了評斷關鍵變數的有效性，本研究將所得的關鍵變數再度放入支撐向量機中，比較其分類指標的優劣來判斷關鍵變數組合的有效性。本研究所得到的智慧型電視的核心功能有：Wifi、2D/3D轉換、運算/儲存能力、體感控制、聲控以及獨立使用者平台等。而在體感控制方面，核心的手勢集應有以下手勢：鼠標式移動選取、選取、連續性調整、取消選取/上一層選單以及開啟雙手手勢控制等。智慧型電視開發商可奠基於以上功能及手勢進行產品開發。
關鍵字：資料採礦、羅吉斯迴歸、粗糙集理論、支撐向量機、購買預測、貝氏產品擴散模型、科技接受模式

關鍵字

科技接受模式 ； 貝氏產品擴散模型 ； 購買預測 ； 支撐向量機 ； 粗糙集理論 ； 羅吉斯迴歸 ； 資料採礦

並列摘要

Televisions have been playing an important role in people’s life. With the rise of smart devices, smart TVs were also provided. With the advantages such as big screen and high calculate and storage capacities, smart TVs have the potential to be control center of internet of things or smart families. During this period, Smart TVs are in the beginning phase of its’ product life and all products in the market are the first or second generation. These smart TV products have big difference in their functions and interfaces. From 2014 Gartner Hype Cycles for technology and marketing, we can see that smart TV industry is in the expectation declining phase. Combining the observation of hype cycle with bass diffusion model, we can find that the present consumers of smart TVs are mostly innovators. In the future, smart TV manufacturers are going to face the challenge from the lagers whose purchase behaviors are mostly base on seeing others having the products instead of their own enthusiasm. So how to create a product which and not only satisfy the innovators but also able to attract lagers are the biggest issue for smart TV manufacturers.
This study uses Logistic regression(LR), rough set theory(RST) and support vector machine(SVM) in function and gesture variables’ analyzation. First, this study analyze all the function and gesture variables by putting them into Logistic regression, rough set theory and support vector machine models. After ensuring the consisting and effectiveness, we use LR and RST to search for key variables. Second, in order to check the key variable combinations’ effectiveness, we put the key variables combination founded in LR and RST into SVM. By analyzing the classification indexes, we can compare the key variable combinations and draw the following conclusions: (1) The core functions for smart TVs are Wifi, 2D/3D transfer, Calculate and storage, Gesture control, Voice control and Individual platform. (2) The key gestures that should be in smart TVs’ gesture control set and Mouse-like move, grab, long grab, CCW(counter clockwise) rotation and Waving two hands. With those core function and gestures, smart TV manufacturers are able to create smart TVs that suit the market better.
Key words：Data Mining, Logistic Regression, Rough Set Theory, Support Vector Machine, Purchase Forecasting, Bass Diffusion Model, Technology Acceptance Model

並列關鍵字

Technology Acceptance Model ； Bass Diffusion Model ； Purchase Forecasting ； Support Vector Machine ； Rough Set Theory ； Logistic Regression ； Data Mining

參考文獻

凃嘉峪, 曾光華, & 何雍慶. (2006). 應用資料採礦技術分析生物科技新產品之市場區隔: 以健康食用油為例. 科技管理學刊, 11(4), 63-98.

張語恬, 朱基銘, 簡戊鑑, 周雨青, 楊燦, 盧瑜芬, ... & 羅慶徽. (2007). 比較三種資料探勘演算法預測子宮頸癌五年存活的外部通用性效能. 臺灣家庭醫學雜誌, 17(4), 222-238.

Ahn, H., Ahn, J. J., Oh, K. J., & Kim, D. H. (2011). Facilitating cross-selling in a mobile telecom market to develop customer classification model based on hybrid data mining techniques. Expert Systems with Applications, 38(5), 5005-5012.

Arimond, G., & Elfessi, A. (2001). A clustering method for categorical data in tourism market segmentation research. Journal of Travel Research, 39(4), 391-397.

Arvola, A., Lähteenmäki, L., & Tuorila, H. (1999). Predicting the intent to purchase unfamiliar and familiar cheeses: the effects of attitudes, expected liking and food neophobia. Appetite, 32(1), 113-126.