English  |  正體中文  |  简体中文  |  Items with full text/Total items : 51929/87065 (60%)
Visitors : 8475625      Online Users : 112
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/102392


    Title: 運用情感運算與類神經網路技術建置服務決策系統
    Other Titles: Building a service decision system by using affective computing and artificial neural network
    Authors: 陳思傑;Chen, Szu-Chieh
    Contributors: 淡江大學資訊管理學系碩士班
    解燕豪;Hsieh, Yen-Hao
    Keywords: 情感運算;服務科學;服務補救;類神經網路;決策樹;Affective computing;Service science;service recovery;artificial neural networks;Decision tree
    Date: 2014
    Issue Date: 2015-05-04 09:54:39 (UTC+8)
    Abstract: 近年來,資訊科技的快速成長,不僅影響了人類的生活環境,更提升產業的工作效率,帶來更為多樣化且便利的生活品質,其中資訊科技的自動化服務,使顧客能夠快速地獲得服務,企業更能利用資訊科技的高運算能力,根據現有資源來提供顧客合適的服務項目,並協助企業制定出最有效的決策。目前企業在提供服務決策時,除了重視顧客的基本需求外,也越來越重視顧客心理的感受,而顧客情緒正會影響企業執行服務決策。情感運算越來越受重視,其概念為資訊科技能夠具備人類之情緒,並於人機互動的過程中,使資訊科技以人性的思維協助企業進行決策,本研究即是認為情緒除了傳達出對於事物的感受程度外,亦能夠表現出其對於服務的評價,因此資訊科技若能夠與人類一樣具有情緒能力,必能制定出更為人性化的決策。目前服務業在台灣產業結構中佔有非常大的比重,於此競爭激烈之市場環境中,企業為了提升其競爭能力,在服務提供上除了滿足顧客的基本需求外,亦期望能夠在服務互動過程中重視顧客之情緒變化,從顧客回饋中創造出服務的體驗價值。因此本研究認為,企業於進行服務決策時,若能夠將情感運算應用於服務決策中,即時地察覺顧客的情緒變化,並加以分析出最適合的補救項目,進而能夠快速且有效地解決顧客的問題。
    本研究利用服務互動設計之概念,透過語音情緒辨識技術、類神經網路與決策樹分類法,建立一服務決策系統,並應用服務補救作為研究情境,驗證情緒在服誤決策提供上之成效。本研究之服務決策系統,運用情感運算之情緒辨識技術,偵測顧客於服務失誤發生時可能會產生的情緒反應,透過倒傳遞式類神經網路進行情緒之分類,並透過C4.5決策樹建立有效之服務補救規則,使系統能夠根據情緒反應與服務失誤情境來提供服務補救項目。為了建置出有效的服務決策系統,本研究利用Matlab R2013a來進行服務決策系統之設計,並設計了三個研究實驗透過四個餐飲服務業之服務失誤狀況,以語音情緒辨識之方式,分析顧客的情緒反應並提供出合適的服務補救項目,並根據顧客的期望補救項目排序來驗證服務決策的有效性。實驗一利用了倒傳遞式類神經網路來建立有效的情緒分類方法,並以試誤法找出適合本研究之類神經網路架構;實驗二透過網路問卷的方式,蒐集顧客對於服務失誤情境下的期望補救排序,其所收集之有效問卷共110筆,並透過C4.5決策樹之演算法,求出適合各情境下之關鍵屬性,用以建置出有效之服務補救決策規則;實驗三根據實驗一與實驗二之實驗結果,建置出服務決策系統,請35名受測者來對服務決策系統進行測試與評估,測試實驗一所建立之情緒分類準確度,以及實驗二所建立之服務決策規則準確度。
    根據本研究實驗之結果,將情感運算結合服務互動設計之概念,所建置之服務決策系統能夠有效地進行情緒辨識,其情緒辨識準確率平均約為72.9%,且能夠提供符合顧客期望之服務補救項目,其服務決策準確率平均約為73.57%,而受測者對於此種服務決策方式,於各情境下所給予的服務體驗評價平均約為87.5%之高度滿意度。因此本研究發現,對企業而言,情緒不僅能夠作為服務決策提供之依據,更能將情緒應用在與顧客的服務互動當中,建立不同於以往的服務傳遞過程,創造顧客嶄新之服務體驗價值。企業在提供服務給顧客時,大致上仍是由顧客主動要求應提供何種服務項目,企業較鮮少去考量顧客獲得服務時的情緒狀況,因此根據本研究所提出之服務決策系統,企業能夠透過語音情緒辨識的方式,即時地了解顧客心理之需求,並提供合適的服務項目,提升顧客對企業之滿意度與忠誠度。本研究主要著重於語音情緒辨識之技術應用,並以服務補救之情境來驗證本服務決策系統之成效,未來期望能夠結合不同的情緒辨識技術,例如臉部表情、肢體動作偵測等等,分類出更多樣化的情緒類型,並透過普及運算之概念,將情緒靈活應用於更多的服務領域中,提供更多元化的服務項目。
    With the development of information technology, information technology not only improves human’s living environments and the quality of life but also increases the productivity of industries. Self-service technology can efficiently to enable customers to acquire proper services. Meanwhile, information technology helps businesses to provide customers with quality services using existing resources and make the right and effective service decisions. Accordingly, businesses have to not only understand customer needs but also pay attention to customer emotions when they make service decisions. Affective computing is an important approach to recognize human’s emotions that information technology has high abilities of revealing and recognizing emotions to businesses to make effective service decisions. The service industry plays an important role in economic activities in Taiwan. Businesses need to create innovative valuable services to customers by understanding customer emotions within service encounters. Consequently, in order to dealing with customer problems, businesses should apply affective computing to service decision processes to recognizing customer emotions and deliver suitable services to customers.
    This study aims to build a service decision system by adopting affective computing, artificial neural networks and decision trees based on the concept of service interaction design. This study uses service recovery as a case to test the service decision performance of the service decision system. Affective computing is to recognize customer emotions when customers encounter service failures. The customer emotions can be analyzed and clustered to positive and negative emotions via artificial neural networks. Then, this study tries to design effective service decision rules based on C4.5 algorithm of decision trees. This study builds the service decision system by using Matlab R2013a tool and defines four scenarios of service failures. Three experiments are conducted to evaluate the service decision system. Experiment 1 is to build an effective approach to classify customer emotions and to explore the proper analysis structure of artificial neural networks. Experiment 2 is to build suitable decision rules for the service decision system by using decision trees and surveying real data through 110 internet subjects. Experiment 3 is to evaluate the performance of the service decision system that 35 subjects are invited to experience the service decision system and respond to the service failures based on the experiment results of experiment 1 and experiment 2.
    The experiment results show that the service decision system is built by combining the concept of service interaction design and affective computing which can have the high performance of customer recognition. The customer recognition rate is about 72.9%, the accuracy rate of service decision for customers is about 73.57 and the customer satisfaction rate for the service recovery is about 87.5%. Hence, according to the result findings, customer emotions can be a clue to enable businesses to make the right service decisions and create innovative service experiences for customers within service interactions. Meanwhile, in order to increase customer satisfaction and loyalty, businesses can effectively understand customer needs and requirements to deliver suitable services when customers encounter service failures. This study only focuses on recognizing customers’ speech to understand their emotions and use the service recovery as a case. Researchers continuously pay attention to the applications of combining face recognizing with action recognizing to cluster more emotion types. Besides, further research can elaborate the proper applications of service fields based on the idea of pervasive computing.
    Appears in Collections:[資訊管理學系暨研究所] 學位論文

    Files in This Item:

    File SizeFormat
    index.html0KbHTML95View/Open

    All items in 機構典藏 are protected by copyright, with all rights reserved.


    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback