110年科技管理學刊第26卷第一期

上層分類: 科技管理學刊 各期學刊目錄

110年科技管理學刊第26卷學刊第一期 民國一一○年六月

Volume 26, Number 1 June 2021(若需下載全文請登入會員)


標題Article大數據平台上基於顧客喜好和社群行為分析之大數據平台上基於顧客喜好和社群行為分析之顧客流失預測Churn Prediction Based on the Analysis of Customers’Preferences and Social Behavior on a Big Data Platform

作者Aurthor:賴錦慧、李維平、伏泳霖 Chin-Hui Lai, Wei-Ping Lee,Yung-Lin FuYung-Lin Fu

中文摘要Chinese Abstract

由於網際網路和電子商務的應用已發展得很完善,顧客有更多的選擇而使得顧由於網際網路和電子商務的應用已發展得很完善,顧客有更多的選擇而使得顧客的忠誠度正逐漸下降。對於企業而言,開發新顧客的成本遠高於保留舊顧客的成本,而顧客的流失也將會對企業造成巨大的影響。因此,企業如何在利用有限資源有效地管理顧客以減少顧客流失對企業的影響,是企業必須面對的重要議題。本研究在一個大數據平台中提出一個應用於社群網路之顧客流失預測的新穎方法,此方法將整合顧客價值分析、顧客喜好、以及社群行為和信任關係之分析等四大因素,再利用分類方法,亦即隨機森林、K-NN 和類神經網路,針對這四個因素進行顧客流失預測。實驗結果顯示本研究所提出的顧客流失預測模型在分析大量數據時,能有較好的預測準確度和計算效能。

關鍵字:資料探勘、顧客流失預測、潛在狄利克雷分配、隨機森林、決策樹、類神經網路

English Abstract
Because of the well-developed Internet and e-commerce applications, customersBecause of the well-developed Internet and e-commerce applications, customershave a lot of choices so that customer loyalty is decreasing. For businesses, thecost of developing new customers is higher than the cost of retaining oldcustomers. Losing customers (users) will cause a huge impact on business.Therefore, how a business uses limited resources to manage customers effectivelyand reduce the impact of customer loss are critical issues. This work will build abig data platform and propose a novel method for predicting customer churn on asocial network, which integrates customer value analysis, customer preferences,and the analysis of social behavior and trust relationship. Then, the classificationmethods, i.e. random forest, K-NN, and neural network, use these four factors tomake the customer churn prediction. The experimental results show that theproposed churn prediction model has better prediction accuracy and computationefficiency in analyzing a huge amount of data.


Keywords: Data Mining, Churn Prediction, Latent Dirichlet Allocation,Data Mining, Churn Prediction, Latent Dirichlet Allocation,Random Forest, Decision Tree, Neural Network


標題Article:資通訊科技世代之虛實通路整合行銷變革探討:資通訊科技世代之虛實通路整合行銷變革探討:以汽車售後服務市場為實證情境A Study of the Online and Offline Integration MarketingStrategy Change: An Empirical Examination of AutomotiveAftermarket

作者Aurthor:郭馨禧、鍾君宇、蔡顯童 Hsin-Si Kuo,Zhong Jun-Yu,Hsien-Tung Tsai

資通訊科技的進步與普及,使得傳統產業之經營模式面臨前所未有的挑戰。本研究主要資通訊科技的進步與普及,使得傳統產業之經營模式面臨前所未有的挑戰。本研究主要探討:汽車零件品牌廠商如何透過社會影響機制,說服其所合作的汽車售後服務通路商,包含零配件代理商與保修廠商,採行網路行銷策略,以滿足現今顧客利用多種不同管道取得多樣化產品及服務之需求。從行銷策略的角度,此一策略變革係為汽車售後服務業者長期競爭優勢之基礎。本研究以社會性影響理論 (Social Influence Theory) 及變革準備 (ChangeReadiness) 之學理觀點,提出新的整合型學理架構,深入探究「汽車零件品牌廠商」影響「汽車售後服務通路商」接受並採用網路行銷策略之過程。本研究結合質性與量化實證之研究方法,與一知名的汽車零件品牌廠商合作,驗證本研究之觀點。實證結果發現:在汽車售後服務之情境下,有效的社會影響機制包含:社會順從(Compliance)、社會認同(Identification)、與社會內化(Internalization)。其次,研究結果顯示:品牌廠商之社會影響機制會具體提升汽車售後服務通路商對於網路行銷之「感知變革適當性」及「感知變革效能」,進而形成其「網路行銷變革意願」及「網路行銷策略採用行為」。本研究也進一步探討「科技準備度」之調節效果,分析結果顯示:對於網路科技持有樂觀偏見之汽車售後服務通路商主管及行銷人員,其網路行銷變革意願與實際採用行為呈現抑制(弱化)之現象。本研究結果除了深化社會影響與行銷策略之議題探討,亦提供經理人若干虛實通路整合行銷實務之建議。
關鍵字:網路行銷策略、社會性影響理論、變革準備、科技準備度、樂觀偏見、虛實通路整網路行銷

English Abstract

The advancement and popularization of information and communicationsThe advancement and popularization of information and communicationstechnology has made traditional industry business models face unprecedentedchallenges. This research mainly explores: how auto parts brand company canutilize social influence mechanisms to persuade the automotive aftermarketchannel members, including parts dealers and service workshops, to adopt Internetmarketing strategies and satisfy the needs of today’s customers who tend to obtaindiversified products and services through multiple channels. From the perspectiveof marketing strategy, this strategic change is the basis of long-term competitiveadvantage. Based on Social Influence Theory and the concept of ChangeReadiness, this research proposes a new integrated theoretical framework and indepthexploration on the influence process of auto parts company towards theautomotive aftermarket channel members to adopt Internet marketing strategy.This research combines qualitative and quantitative empirical research methods,and cooperates with a well-known auto parts brand manufacturer to verify thetheoretical model of this research. The empirical results indicate that the effectivesocial influence mechanisms of auto parts company, including social compliance,social identification, and social internalization, will enhance the “perceivedchange appropriateness” and “perceived change efficacy” of automotiveaftermarket channel members, and lead to their intention and adoption of Internetmarketing strategy. Finally, this research explores the moderating effect of“technology readiness”. The result shows that the individuals with optimism biastowards Internet technology tend to overestimate their actual performance inadopting Internet marketing strategies. In addition to deepening the research onsocial influence and social marketing, the results of this research also providemanagers with suggestions on management practices on the integration of onlineand offline marketing.


Keywords: Internet Marketing Strategy, Social Influence Theory, ChangeInternet Marketing Strategy, Social Influence Theory, ChangeReadiness, Technology Readiness, Optimism Bias, Integration ofOnline and Offline Marketing


 

標題Article捲土重來!募資失敗再上市之意會活動歷程探討捲土重來!募資失敗再上市之意會活動歷程探討Bounce Back! An Exploratory Study of SensemakingProcess in Crowdfunding Failure and Relaunching

作者Aurthor吳曉君陳悅琴、陳品綺 Shiao-Jun Wu, Yueh-Chin Chen, Pin-Chi Chen

中文摘要Chinese Abstract

群眾募資是近幾年來新創團隊取得資金、低成本行銷和測試消費偏好群眾募資是近幾年來新創團隊取得資金、低成本行銷和測試消費偏好的管道之一,但募資專案提案者卻承受著近五成的失敗率。過去研究多探討如何在募資平台募資成功、或從消費者端分析消費者贊助偏好,但卻鮮少研究探討這些高達五成失敗的募資案何去何從。本研究訪談六位募資失敗再上市之提案者,探討經歷募資失敗之提案者如何反思失敗經驗、如何在自我效能趨動下學習新技能與優化產品或服務、如何意會環境需求而調適營運活動和展現策略行動。本研究有助於了解失敗募資提案者捲土重來之意會活動歷程,提供募資提案者轉化失敗經驗為資產、而非負債之觀點。

關鍵字:群眾募資、失敗經驗、自我效能、意會活動、新創事業

English Abstract

Crowdfunding is one of the conduits for start-up team to obtain capital, low-costCrowdfunding is one of the conduits for start-up team to obtain capital, low-costmarketing and test consumer preferences in recent years, but the crowdfundingproposers bear a failure rate of up to 50%. Past studies have explored how to raisemoney successfully on a crowdfunding platform or analyze consumer sponsorshippreferences from the consumer side, but few have explored where these failedcrowdfunding cases can go. The research interviewed with six sponsors who failedto raise money and went back to market. It explored how those who failed to raisemoney could reflect on failures, learn new skills and optimize products or serviceswhen driven by self-efficacy, adapt operations to environmental needs, anddemonstrate strategic actions. This study will help to understand the sensemakingprocess of the failed crowdfunding and comeback, and provide the view thatcrowdfunding proponents convert failed experiences into assets rather thanliabilities.

Keywords: crowdfunding, failure experience, self-efficacy, sensemakingcrowdfunding, failure experience, self-efficacy, sensemakingprocess


標題Article基於模組化模糊推理系統處理大數據問題基於模組化模糊推理系統處理大數據問題Processing Big Data Problems using Modular FuzzyInference Systems

作者Aurthor李昕潔、劉學承﹑曾敬勳 Shin-Jye Lee, Hsueh-Cheng Liu, Ching-Hsun Tseng

中文摘要Chinese Abstract

訊息技術近幾十年來的飛速發展導致每天生成和交換的數字數據量不斷增加,訊息技術近幾十年來的飛速發展導致每天生成和交換的數字數據量不斷增加,有效地從海量大數據中獲取訊息為公司帶來競爭優勢已經成為一個日益嚴重的問題,本研究通過建立模塊化的模糊推理系統,提出了一種基於大數據的決策支持策略,該系統主要由模糊系統識別,k 均值聚類算法和自適應網絡模糊推理系統(ANFIS)組成,基本上,ANFIS 被設計為基於模糊If-Then 規則具有優化方案的推理系統,特別是在運籌學領域,它的學習能力為決策支持系統帶來了積極的表現且常被用於支持各種決策問題。


關鍵字:

科技管理、模糊系統識別

English Abstract

Over recent decades, the rapid progress of information technology has led to anOver recent decades, the rapid progress of information technology has led to anever-increasing amount of digital data being generated and exchanged every day.Therefore, effectively acquiring the information from this vast amount of big datato give a company a competitive advantage has become an increasing issue. Thisarticle proposes a decision support strategy based on big data by establishing amodular fuzzy inference system, and the proposed system primarily consists offuzzy system identification, k-means clustering algorithm, and adaptive networkbasedfuzzy inference system (ANFIS). ANFIS is designed as a Fuzzy If-Thenrules-based inference system with the optimization scheme. In particular, itslearning ability brings positive performance to the decision support system, and itis usually applied to support a variety of decision-making issues in operationalresearch.

Keywords: Technology Management, Fuzzy System Identification

列印