市場趨勢導向機會發掘之服務價值創新 - 政大學術集成

全文

(1)國立政治大學資訊管理學系. 碩士學位論文指導教授:苑守慈博士. 立. 政治大. ‧. ‧ 國. 學. 市場趨勢導向機會發掘之服務價值創新. n. al. er. io. sit. y. Nat. Market-Oriented Chance Discovery toward Service Value Creation. Ch. engchi. i n U. v. 研究生:沈品勳中華民國 103 年 6 月.

(2) Market-Oriented Chance Discovery toward Service Value Creation 市場趨勢導向機會發掘之服務價值創新. by. Pin-Hsun Shen. A Dissertation Submitted in Total Fulfillment of. the Requirement for the Degree of. 立. 政治大 Master of Science. In. ‧ 國. 學 Management information Systems. ‧ y. Nat. n. al. er. io. sit. Supervisor: Soe-Tysr Yuan, Professor, MIS, NCCU. Ch. engchi. i n U. v. Department of Management Information Systems. NATIONAL CHENGCHI UNIVERSITY. June 2014. © Pin-Hsun Shen 2014.

(3) 致謝. 首先誠摯地感謝指導教授苑守慈老師，在碩士的這些年細心地教導，讓我不斷地吸收服務科學領域專業知識並整合至技術領域。在撰寫論文的過程中，也不斷地討論並指點迷津，訓練我在完成任務上的嚴謹度及標準，讓我在課內和課外中獲益匪淺。老師對多專案的管理和清楚思維也將會是我一輩子追求的境界。. 立. 政治大. ‧ 國. 學. 三年裡的日子，不斷地想起實驗室裡和同學張薰方、謝沛剛、梁. ‧. 煜銜的生活點滴，不論是學術上的討論、課外的閒扯、半夜睡在實. sit. y. Nat. 驗室沙發、趕論文和寫系統進度壓力太大的苦日子等。雖然碩士第. er. io. 三年自己去了外面闖闖，但還是很感謝和時常回憶前兩年眾學長. n. a. v. l C 姐、同學、學弟妹的陪伴，讓我的政大研究生活有哭有笑。 ni. hengchi U. 此外，女朋友是我的一大支柱，在我學位考試時背後的支持更是我疲累不堪時拾起勇氣的動力。最後，謹以此文獻給我摯愛的雙親，口試通過的一刻，最開心的是聽到父親在電話中的愉悅與興奮，沒有甚麼事情比的上父母的笑容來的更重要。品勳於民國一 O 三年六月十日 i.

(4) Abstract. Market orientation refers to the business philosophy that focuses on identifying and meeting the stated or hidden needs of customers through various approaches of intelligence gathering activities. To this end, two important capabilities, market sensing and market relating, are to enable business to formulate, examine, modify, renovate and redefine their market views and investigate among all players in the market. This study focuses on how to exploit information system as a service to facilitate the acts of the. 政治大 the text mining technique 立 and the chance discovery theory. In addition, this service. market sensing and market relating capabilities in terms of information technologies of. ‧ 國. 學. would help various businesses seek new, necessary and related resources or fellow companies for each other to cooperate to form a complex service product from different. ‧. markets. In other words, cooperating in the open relationship, different companies. sit. y. Nat. attain more possibility to mutually expose their features, product or service, which. n. al. er. io. achieves a marketing for those companies in different markets as well, and better market. i n U. v. orientation is also avail business to make better complex product or service for niche. Ch. engchi. market to get competitive advantages.. Keywords: SMEs, Operant Resource, Sensing capability, Market Relating capability, text mining, Chance discovery. ii.

(5) Table of Content Abstract ........................................................................................................................... i Figure Content ............................................................................................................... v Table Content ............................................................................................................... vii CHAPTER 1 INTRODUCTION ................................................................................... 1 1.1 Background and Motivation ............................................................................ 1 1.2 Research Question ........................................................................................... 3 1.3 Research Method ............................................................................................. 5 1.4 Purpose and Contribution ................................................................................ 5 1.5 Content Organization ....................................................................................... 7 CHAPTER 2 LITERATURE REVIEW....................................................................... 10. 治政 2.1 Market Orientation ......................................................................................... 10 大 2.2 Operant Resource立 Capability ......................................................................... 12 ‧. ‧ 國. 學. 2.3 Market Sensing Capability............................................................................. 13 2.4 Term processing technique............................................................................. 16 2.4.1 Word Segmentation: ............................................................................ 16. n. al. er. io. sit. y. Nat. 2.4.2 Term weighting schemes ..................................................................... 16 2.5 Vector Space Model ....................................................................................... 18 2.5.1 Term-Document Matrix ...................................................................... 19 2.5.2 Similarity Calculation ......................................................................... 19 2.6 Document Clustering: .................................................................................... 20 2.7 News Event tracking and detecting................................................................ 21 2.8 Story, Event and Topic ................................................................................... 23 2.9 Market Relating Capability ............................................................................ 23 2.10 Chance Discovery ........................................................................................ 26. Ch. engchi. i n U. v. 2.10.1. KeyGraph ......................................................................................... 27 2.10.2. Data Crystallization ......................................................................... 28 CHAPTER 3 PROSPERITY TAIWAN INTEGRATED PROJECT ........................... 30 3.1 Taiwan’s Economic Strengths and Current Economic Background .............. 30 3.2 The “Prosperity” in Prosperity Taiwan .......................................................... 31 3.3 Vision of Prosperity Taiwan ........................................................................... 31 3.4 Culture, Arts and Creativity as an Example ................................................... 31 3.5 Intelligent Service Machines to aid Economic Transformation ..................... 32 3.5.1 The V+ Platform ................................................................................. 33 3.5.2 Application Scenario of V+ platform .................................................. 35 3.5.3 Application Scenario of Sourcing+ ..................................................... 36 iii.

(6) CHAPTER 4 METHODOLOGY ................................................................................ 41 4.1 Conceptual Model .......................................................................................... 41 4.2 System Architecture ....................................................................................... 45 4.3 Market Sensing .............................................................................................. 46 4.3.1 Story Sensing Module ......................................................................... 47 4.3.2 Event Detecting and Tracking Module ............................................... 52 4.4 Market Relating ............................................................................................. 54 4.4.1 Chance Discovery Module .................................................................. 55 4.4.2 Existing Chance Module ..................................................................... 63 4.4.3 New Chance Module................................................................................... 65 CHAPTER 5 APPLICATION SCENARIO ................................................................. 68 5.1 An Application Scenario .......................................................................... 68 5.2 Service Blueprint ..................................................................................... 74 CHAPTER 6 EVALUATION ...................................................................................... 76. 政治大 6.1 Propositions.................................................................................................... 77 立 6.2 Assumptions ................................................................................................... 79. ‧ 國. 學. 6.3 Experimental Design and Results .................................................................. 80 6.3.1 Experiments Design ............................................................................ 80. ‧. 6.3.2 Propositions Verification and Results ................................................. 93 6.4 Discussion .................................................................................................... 120 6.4.1 Proposition ........................................................................................ 120 6.4.2 Other Discussion ............................................................................... 123 CHAPTER 7 CONCLUSION.................................................................................... 131. er. io. sit. y. Nat. al. n. 7.1 Research Contribution ................................................................................. 133 7.2 Limitation ..................................................................................................... 135 7.3 Future Work ................................................................................................. 137 REFERENCES .............................................................................................................. a. Ch. engchi. iv. i n U. v.

(7) Figure Content Figure1-1. A conceptual framework for developing complex, business-to-business products .......................................................................................................................... 4 Figure1-2. Henver’s Design Science Research Framework .......................................... 8 Figure 1-3. Conceptual framework to examine differences in MO ............................. 11 Figure 2-1. Relationship between market sensing and market orientation .................. 14 Figure 2-2. Term-Document Matrix (Salton and Gill, 1983) ....................................... 19 Figure 2-3.Topic Detecting and Tracking (Panka, 1999) ............................................. 22 Figure 2-4. Three elements of market relating capability(Day, 1999) ......................... 24 Figure 2-5. Data of a cluster ........................................................................................ 27 Figure 2-6. an Example of KeyGraph Method ............................................................ 28 Figure 3-1. A simple service machine is a hybrid of STS and SSME ......................... 33 Figure 3-2. The V+ Platform........................................................................................ 33. 立. 政治大. ‧. ‧ 國. 學. Figure 3-3. The Service Process of Value+ Engine ..................................................... 37 Figure 3-4. Different Type of Market Intelligence....................................................... 38 Figure 3-5. Instances of event ...................................................................................... 38. n. al. er. io. sit. y. Nat. Figure 3-6. Cooperating condition of 2nd stage .......................................................... 40 Figure 4-3. Insight after Semantic Analysis ................................................................. 48 Figure 4-4. Combination of keywords for google query ............................................. 48 Figure 4-5. The process of new story emerging........................................................... 52 Figure 4-6. The process of Multi-Document Abstract Generating System ................. 54 Figure 4-7. Processing Keygraph by Polaris................................................................ 56 Figure 4-8. Pre-processing of raw data & Reanalyzed map after filtering .................. 57 Figure 4-9. Important concept in event map ................................................................ 58 Figure 4-10. a clustering result of Pictorial KeyGraph for the data from event/ topic. Ch. engchi. i n U. v. database ........................................................................................................................ 59 Figure 4-11. Event map with dummy nodes ................................................................ 61 Figure4-12. Pictogram of Dummy 2-42 (drawings, charts and text of Patent)............ 61 Figure4-13. Pasting Pictorial on Event Map ................................................................ 62 Figure4-14. Four clusters with hidden events with Pictorial on event Map ................ 62 Figure 4-16 comparing the searching time through Google Trends ............................ 64 Figure4-17. Network flow of cbav.com.tw through Alexa .......................................... 64 Figure 4-18. Possible Emerging Scenario .................................................................... 66 Figure4-19. New scenario after evaluation .................................................................. 66 Figure 5-1. The process of 1st system stage ................................................................. 69 Figure 5-2. Insight analysis .......................................................................................... 69 v.

(8) Figure 5-3. Market intelligence with importance, newness and relevance degree ...... 70 Figure 5-4. Related event in Relevance Degree........................................................... 71 Figure 5-5. Market intelligence with importance, newness and relevance degree ...... 71 Figure 5-6. Final Event Map on Polaris ....................................................................... 73 Figure 5-7. Uploading event map and new chances .................................................... 73 Figure 5-8. The Blueprint of 1st system stage(market sensing capability) .................. 74 Figure 5-9. The Blueprint of 2nd system Stage(market relating capability) ................. 75 Figure 6-1. Experimental Design ................................................................................. 81 Figure 6-2. The proposition testing of P1 .................................................................... 93 Figure 6-3. Mathematical Symbols of P1 .................................................................... 94 Figure 6-4. Value of ta ................................................................................................ 98 Figure 6-5. The proposition testing of P2 .................................................................. 111 Figure 6-6. Mathematical symbols of P2 ................................................................... 111 Figure 6-7. one tail Z table ......................................................................................... 112 Figure 6-8. The proposition testing of P3 .................................................................. 119. 立. 政治大. ‧ 國. 學. Figure 6-9. the Line Chart of Three Market Sensing Degrees of Title Rank ............. 125 Figure 6-10. the Line Chart of Three Market Relating Characteristics of Title Rank .................................................................................................................................... 126. ‧. Figure 6-11. the Bar Chart of Three Market Sensing Degrees of Motivation ........... 127 .................................................................................................................................... 129 Figure 6-12. the Bar Chart of Three Market Sensing Degrees of Insight Type ......... 129 Figure 7-1. Implicit process of IT artifact .................................................................. 134. n. er. io. sit. y. Nat. al. Ch. engchi. vi. i n U. v.

(9) Table Content Table 2-1. Comparison of existing methods of market sensing ................................... 15 Table 2-2. Local Weight in an article ........................................................................... 17 Table2-3. Common Formula of Global Weight ........................................................... 17 Table 2-4. Common Similarity Calculation ................................................................. 19 Table 2-5. Comparison of existing methods of market relating ................................... 25 Table 4-1. Three websites to text mining ..................................................................... 49 Table 4-2. Analysis of newness, importance, similarity of degree .............................. 51 Table 4-3. Process of data crystallization in Pictogram method .................................. 60 Table 6-1. Experimental Subject Information .............................................................. 85 Table 6-2. Evaluation Level and Effect of Importance, Newness and Relevance Degree .......................................................................................................................... 86 Table 6-3. Evaluation of the three attributes of new chances generated from sourcing+. 立. 政治大. ‧. ‧ 國. 學. ...................................................................................................................................... 88 Table 6-4. Variables of Macro Evaluation ................................................................... 90 Table 6-5. 5 Evaluation Points of Proposition 3 .......................................................... 91. n. al. er. io. sit. y. Nat. Table 6-6. Statistic Data of proposition 1 .................................................................... 95 Table 6-7. Statistic Descriptive statistics ..................................................................... 96 Table 6-8. Formula and interpretation of one-sample T test ........................................ 97 Table 6-9. One-Sample T Test of importance, newness and relevance degree ............ 97 Table 6-10. Field Data Encoding (1) .......................................................................... 102 Table 6-11. Field Data Encoding (2) .......................................................................... 104 Table 6-12. Field Data Encoding (3) .......................................................................... 107 Table 6-13. Statistic Data of proposition 2 ................................................................ 109 Table 6-14 . Comparison between before using Sourcing+ and after using Sourcing+. Ch. engchi. i n U. v. .................................................................................................................................... 115 Table 6-15. Statistic Data of proposition 3 ................................................................ 117 Table 6-16. the Aanalysis of Title Rank ..................................................................... 123 Table 6-17. Average Scores of Three Market Sensing Degrees of Title Rank ........... 124 Table 6-18. Average Scores of Three Market Relating Characteristics of Title Rank124 Table 6-19. the Aanalysis of Motivation .................................................................... 126 Table 6-20. Average Scores of Three Market Sensing Degrees of Motivation .......... 127 Table 6-21. the Aanalysis of Insight Type .................................................................. 128 Table 6-22. Average Scores of Three Market Sensing Degrees of Insight Type ....... 128 Table 6-23. the Comparision between the two Extreme values ................................. 130 Table 7-1 . 30 Interviewers’ Data for system future application ................................ 139 vii.

(10) CHAPTER 1 INTRODUCTION. 1.1 Background and Motivation With the growth of global economy liberalization and internationalization, small and medium enterprises (SMEs) are up against more pressure to compete with domestic and overseas markets. In a new dynamic competition environment, traditional subcontractor industry has gradually lost competitive advantages due to scarce of capital-collecting. 政治大. ability, margin, innovation and talents.. 立. In numerous markets, larger companies plays the main roles because of advantages. ‧ 國. 學. like better economies of scale, bargaining power with suppliers/distributors, brand name reputation, experience of curve effects and ahead of monopoly to set prices. ‧. (Fiegenbaum and Karnani, 1991). On the other hand, SMEs often are inferior to larger. y. Nat. io. sit. companies. Nevertheless, SMEs have other advantages such as more innovative. n. al. er. especially in the early period of life cycle (Acs and Audretsch, 1987,2002), more. i n U. v. customer contact (Coviello et al., 2000), greater propensity for action and more output. Ch. engchi. flexibility. Another advantage is SMEs can integrate elements such as information processing, knowledge, and responsiveness into a unique strategic resource (Olavarrieta and Friedmann, 2008). Therefore, SMEs should utilize their advantages to reverse weakness in the market. In Resource-based view (Wernerfelt, 1984), special resource (tangible or intangible) of company is in accordance with competitive advantages. Therefore, combining different resources to make innovative product/service in RBV is the key point to Blue Ocean Strategy. In addition, companies should emphasize operant resources (competences) to get competitive advantages for innovative solutions 1.

(11) (Hafeez and Essmai, 2007). While operant resources exerted across different companies (i.e., co-developing and combining operant resources cross industry-domain businesses), complex product is often involved. Complex product means “systemic, having multiple interactions, and non-decomposable product” (Singh, 1997), which is often involved while operant resources which act other resources contrary to operand resources (Vargo and Lusch, 2004) are exerted across different companies. In a firm’s position, to develop a successful business to business (B2B) complex product by systematically creating,. 政治大 and business involve two important factors which are analyzed from effectiveness of 立. digging out potential opportunities and useful operant resources from market customer. market variables view and achievement capability view.. ‧ 國. 學. 1. Market Orientation:. ‧. First of the important factor related to effectiveness of market variables is market. sit. y. Nat. orientation (MO) (Madhavaram and Appan, 2010). MO relates to organization culture. io. er. emphasizing customer, competitor coordination and inter-functional coordination as three successful keys (Kohli and Jaworski, 1990; Narver and Slater,1990).. al. n. v i n MO has distinct difference C between SME and big company. Raju, Lonial and Crum hengchi U. (2011) referred to a conceptual model which emphasizes five aspects such as. Organizational Structure, Organizational Culture, Market Orientation, Environment Moderators, Mediators and Performance to measure mutual effect included many studies to distinguish MO’s effectiveness between the two. In the five aspects, the study of Market Orientation is emphasized most because of its influential factors like customer orientation, competitor orientation, responsiveness and interfunctional coordination being interactions between the stakeholder of business and market. As a result, compared to big company, SMEs have certain significance such as (1) Ability to focus on the market is the successfully key word for SME (Ghosh et al., 2001) 2.

(12) (2) Competing in narrow market niche facilitates SME’s intelligence generation (Chen and Hambrick , 1995). (3) Linkage between MO and product innovation would advance SME’s performance (Salavou, 2002). (4) MO effect SME’s innovative capability ( Low et al, 2007). 2. Operant Resource Capability: Several studies pointed out the importance of knowledge, capabilities and competences in complex product (Madhavaram and Appan, 2010; Hofer and Halman, 2005; Singh,1997 et al.). All the concepts of capability, dynamic capability are regarded. 政治大 operant resource capability is similar to intellectual capital knowledge and knowing 立. as operant resource (Constantin and Lusch, 1994). The competitive advantages of. capability of social entirety such as organization, intellectual community and so on. ‧ 國. 學. (Nahapiet and Ghoshal, 1998). Altogether, operant resource should be proactively. ‧. sought, developed integrated by firms, which is along with potential predominance.. Nat. er. io. sit. y. 1.2 Research Question. According to the beginning of our research background and motivation, our research. al. n. v i n Ctwo question direction will follow the of complex product: h efactors ngchi U 1. Market Orientation:. Due to the predominance between MO and SME, potential enterprises/resources have more opportunities to be found and collaborate with each other than the previous traditional methods. The process of traditional complex products is usually confined in a closed and conventional social relationship or working by their own observation and ideas, and the opportunity to innovate and co-create with other companies and distinguish new tendency from market is few. Therefore, the first research question we offered here is “What practicable mechanism can enhance SMEs’ Market Sensing and Market Relating Capability?” 3.

(13) 2. Operant Resource Capability: Operant resources are regarded as an influencing factor in analyzing complex product. In Figure 1-1, there is a conceptual model (as Figure 1-1) analyzing the influencing factors of complex product (Sreedhar and Radha, 2010).. Internal Social Capital - Structural Dimension - Related Dimension - Cognitive Dimension. Operant Resource -Market Sensing Capability Market Relating Capability Market Response Capability Product Developing capability Customization Capability Co-creation Capability. External Social Capital - Structural Dimension - Related Dimension - Cognitive Dimension. 立. Outcomes Complex product development Firm Performance. 政治大. ‧ 國. 學. Figure1-1. A conceptual framework for developing complex, business-to-business products (Madhavaram, 2009). ‧. As we can see, the conceptual model revealed two capabilities, market sensing and market relating capabilities, which enable business to formulate, examine, modify,. y. Nat. io. sit. renovate and redefine their market views (Anderson & Narus, 2007) and investigate. n. al. er. relationship among all players in the market. The two capabilities are affected by. Ch. i n U. v. internal and external social capitals, and, the two capitals also influence other. engchi. capabilities like market response capability to offering or developing product/service to cope with market dynamics, product development capability to facilitate process of linking with customer and firm’s technology, co-creation capability to cooperate with supplier and customer and customization capability to customize product in more efficient or effective way. As a result, better operant resource capability by reinforcing whole six capabilities will lead better outcomes of complex product and firm’s performance. Based on above illustration, the question we propose here is “Do these two capabilities enable SMEs to discover market trend and new chance to create service value?” 4.

(14) 1.3 Research Method Based on the advantages market sensing and market relating capability, this thesis dedicates on enhancing the two capabilities. The following brief explanations of the two capabilities will give a basic view: 1. Market sensing capability: Sensing market orientation (as known as culture-oriented) makes enterprises to sense the customer’s preference, popular resource and reputation assets in different markets. 政治大 resources to innovate (Olavarrieta & Friedman, 1999). 立. to analyze market turbulence, technological change, competitive intensity and new. 2. Market relating capability:. ‧ 國. 學. The three elements of market relating capabilities are orientation, knowledge and skills,. ‧. and integration and alignment of processes (Day 1999). Market relating capability helps. sit. y. Nat. participators (usually referring to R&D staff or sales manager) find a new chance from. io. organizations.. er. existing events in considering mindset, values, and norms inside or among. al. n. v i n C h system to facilitate We will develop an information the two capabilities. For engchi U. combining text mining tool, the system will enhance market sensing capability. On the other hand, for combining chance discovery theory (Ohsawa, 2003), the system will be good in terms of market relating capability. Finally, the system will be a bridge connecting user’s relationship as well as market orientation to achieve operant resource integration for service value creation. 1.4 Purpose and Contribution However, in academic research, the observation of MO is almost performed in qualitative research, the past studies usually only focus on examining the relational 5.

(15) effectiveness among performance, environment, business and consumer’s variables in market, etc. to point out the importance of MO. For instance, numerous studies refer the positive or negative influencing relationship among responsiveness, customer, strategic innovativeness, customer, competitor, organization structure and culture (Nakata et al., 2009; Raju et al., 2011) . For example, the greater levels of organization’s structural level are generally arrest MO within organizations (Jaworski & Kohli, 1993). On the other hand, the four aspects of organizational culture organizational learning, market focus, entrepreneurial proclivity, and quality context are quite relevant to SMEs. 政治大 & Leeflang, 2009; Miller, 1983; Matsuno et al., 2002; Benson et al., 1991). The two 立. and provide SMEs advantages over larger organization. (Hurley & Hult ,1998; Verhoef. factors of MO culture innovation and quality practices can lead to organizational. ‧ 國. 學. performance (Zhou et al., 2008). Finally, the four major moderators market turbulence,. ‧. technological change, competitive intensity and market growth also influence MO-. sit. y. Nat. performance (Jaworski and Kohli,1993; Olavarrieta and Friedmann, 2008; Slater and. io. er. Narver, 1994). In contrast, little literatures emphasize how to realize MO, so our new method dedicates to realize MO by enforcing market sensing and market relating. al. n. v i n C h product is valuable capabilities to make market innovative in nowadays academic engchi U domain.. In order to effectively and efficiently create complex product with optimizing the two factors, information technology (IT) would be a means to conduce SMEs to accelerate production process and marketing, intensify the connection among business goal, ability and market tendency.. and enhance MO capability in domestic or. international market. Therefore, our new method is developing a B2B2C information system helps various businesses seek new, necessary and related resources or fellow companies for each other to cooperate to form a complex product from different markets. On the other 6.

(16) hand, we will obtain the goals resulting positive influence of market orientation, business performance and successful complex product through enhancing market sensing as well as market relating capability. In short, this thesis is dedicated on the research questions about “A new factual method to realize MO”, “Is the method can really help business to find new chance and successful resource combination to make an innovative product/service to satisfy target market?”, and “What are the advantages of the method to industry and academic?” 1.5 Content Organization. 政治大 Basically, the research of the thesis is organized by the principle of Henver’s design 立. science research framework (Henver, March, Park and Ram, 2004). In Figure 1-2,. ‧ 國. 學. Henver’s design science research framework illustrates the relationship among the. ‧. research’s essential factors Environment, Knowledge Base and IS Research. Problem. sit. y. Nat. space is defined in factor Environment, applying theory foundation as well as. io. er. methodology in knowledge base. To pursue the solution of organization’s or object’s difficulty, IT artifact is developed.. n. al. Ch. engchi. 7. i n U. v.

(17) 立. 政治大. ‧ 國. 學. Figure1-2. Henver’s Design Science Research Framework. ‧. In other words, important factors such as environment, IS Research, knowledge. y. sit. io. n. al. er. introduction:. Nat. base, relevance and Rigor will be elaborated in the research chapters as bellow in briefly. . Chapter 1 – Introduction. Ch. engchi. i n U. v. To provide an overall understanding of research background, motivation, question and purpose, this chapter expatiates on present environment and difficulties to define problem space, business needs and opportunities. . Chapter 2 – Literature Review Literature Review delineates theory foundation as well as methodology in knowledge base. The previous study includes domain of technology, management and design. Through raising these studies, chapter 2 gives the backgrounds of concept composed to problem space in chapter1, IT artifacts design in chapter 3 and our methodology in chapter 4. 8.

(18) . Chapter 3 – Prosperity Taiwan Integrated Project Application In this part, we will introduce the whole picture of research project “Prosperity Taiwan”. From high level to bottom level, each component of this project will be pointed out after understanding the entire concept, architecture and process. The system Sourcing+ constructed according to this thesis just belongs to the underlying components of the project.. . Chapter 4 – Methodology In order to establish IT artifacts to correspond with environment problem, this. 政治大 simulate how to enhance firm’s market sensing and relating capability. Each 立. chapter introduces conceptual model and system architecture of our research to. component, relationship of components and methodology of the system will be. ‧ 國. 學. depicted in detailed. Therefore, proposing these IT artifacts represents IS research. Chapter 5 – Application Scenario. sit. y. Nat. . ‧. in design science.. io. er. Chapter 5 simulates the whole service process by applying project Prosperity Taiwan. In this application scenario, a certain SME’s problem will be assisted from. al. n. v i n each stage of Prosperity C Taiwan. This chapterUwill illustrate the scenario of hengchi. Sourcing+, presenting amply service flow and output advantages for the SME. . Chapter 6 – Evaluation The evaluation method is based on one of design science’s guideline to measure the quality, effectiveness of IT artifacts. In this chapter, assumptions and propositions are used to witness to research questions in chapter 1.. . Chapter 7 – Conclusion The conclusion, research contribution, research limitation and future work are mentioned in chapter 7. The excess discovery turns into knowledge base in design science research.. 9.

(19) CHAPTER 2 LITERATURE REVIEW. In the beginning of chapter 2,we will review the previous study from market orientation which is the first factor affecting complex product. Secondly, we accentuate the importance of operant resource capability. Thirdly, market sensing capability will be elaborated. Fourthly, our market sensing capability method involves lots of text mining concept like word segmentation, term weighting scheme, vector space model, term document matrix and similarity calculation and document clustering, topic detecting. 政治大 this part. Finally, we focus立 on adapting chance discovery to realize market relating. and tracking and multiple document abstract, so related study will also be included in. ‧ 國. 學. capability, and basic concept of chance discovery will also be introduced here.. ‧. 2.1 Market Orientation. sit. y. Nat. MO is recognized as a key component of organization in market study. MO is composed. n. al. er. io. of factor like customer orientation, competitor orientation, and inter-functional. i n U. v. coordination (Narver and Slater, 1990) and intelligence generation, intelligence. Ch. engchi. dissemination, and organization-wide responsiveness (Kohli and Jaworski, 1990). But customer orientation, competitor orientation, responsiveness and inter-functional coordination are used in integrated way (Hult et al., 2005). Figure 1-2 is a conceptual model to illustrate important antecedents (organizational structure, organization culture) and consequence (environment moderators, mediators and performance) of MO. The variables in SMEs’ environment and performance are posed by reviewing numerous previous studies (Jaworski & Kohli, 1993; Hurley & Hult ,1998; Verhoef & Leeflang, 2009; Miller, 1983; Matsuno et al., 2002; Benson et al., 1991; Zhou et al., 2008; Olavarrieta and Friedmann, 2008; Slater and Narver, 1994), proving and. 10.

(20) connecting them (Raju et al., 2011). The most important thing is that market orientation positively influences innovation and quality practice of mediators, then affects product and market development of performance, customer loyalty and financial performance.. 立. 政治大. ‧ 國. 學. Figure 1-3. Conceptual framework to examine differences in MO. ‧. (Raju et al., 2011). y. Nat. In order to being market orientation, the two approaches “a market-driven approach”. er. io. sit. and “driving market approach” are proposed (Jaworski & Kohli & Sahay 2000): 1. Market driven: MO of this method is to comprehend and interact actor’s behavior. al. n. v i n C hFor instance, a lotUof apps are developed in accord and attribution in an assured market. engchi with the application of Facebook, because Facebook is the most popular social network website in the market. 2. Driving market: in contrast to market driven, MO affects market structure or actors behaviors in enforcing firm’s competitive position. For example, Google glasses will impact lots of industries in the future, there will be more apps designed to match Google glasses in multiple domains. The ability of Market-driven firms is more excellent in sensing events and trends of market than their competitors. The firm predicts and designs a decision more accurately to attract their customers, enhance channel relationship and fight opponents in time 11.

(21) limitation. MO is emphasized in this section in order to point out the final goal of the research. 2.2 Operant Resource Capability Constantin and Lusch (1994) have made an inception of operant resources concept. The concepts regarded as “operant resource” involve competences, capabilities and dynamic capabilities, therefore “operant resource” here more related to operant resource “capabilities” (not merely represent operant resource). Madhavaram and. 政治大 operant resource to obtain potential competitive advantages. 立. Appan (2010) gave emphasis to proactively search, evolve and manage firm’s own. Madhavaram and Appan (2010) also referred the conceptual model to examine the. ‧ 國. 學. mutual relationship among Internal Social Capital, External Social Capital (Nahapiet. ‧. and Ghoshal , 1998; Tsai and Ghoshal,1998), Operant Resource and Outcome by. sit. y. Nat. integrating lots of past studies (as Figure 1-1), that is, the detailed sub-factors such as. io. er. structural dimension, relational dimension and cognitive dimension of internal social capital in Figure 1-1 are verified that the relationship among other sub-factors is. al. n. v i n Cinhacademic study ofUinternal, external social capital, existent. The mutual relationships engchi. operant resource and outcomes are verified (Bagozzi, 1975; Lusch et al., 1992; Atuahene-Gima, 2002; Hafeez and Essmail, 2007, Hunt 2000, Tsai and Ghoshal, 1998). The six operant resource capabilities provided by Social Capital with sharing knowledge and capability each others are Market Sensing Capability, Market Relating Capability,. Market. Response. Capability,. Product. Development. Capability,. Customization Capability and Co-Creation Capability, they will positively influence the outcomes of Complex product development and Firm Performance. Because of the importance of Market Sensing and Market Relating Capability we mentioned in the first character, our research method is dedicated to enhance the two 12.

(22) operant resource capabilities to reinforce other four capabilities. Therefore, the following literatures will demonstrate related studies of Market Sensing Capability and Market Relating Capability. 2.3 Market Sensing Capability The concept of market sensing is defined as learning something current and prospective customers and competition in the market. Marker Sensing makes company to formulate test, revise, update and improve their original market view. Market sensing can also be. 政治大 customer value and gaining customer feedback. In pursuing the capability, the firm 立. separated to various facets like defining the market, monitoring competition, assessing. should (Anderson & Narus 2007).. ‧ 國. 學. Foley and Fahy (2004) proposed a study of proving the positive relationship. ‧. between market sensing capability and market orientation. And four important. sit. y. Nat. components such as learning organization, organizational system, market information. io. er. and organizational communication in market sensing capability are generalized in Figure 2-1. If successfully enhancing the factors of each component, market sensing. al. n. v i n C hmarket orientationUeven to business performance. capability will positively influence engchi. 13.

(23) 立. 政治大. ‧. ‧ 國. 學. io. sit. y. Nat. Figure 2-1. Relationship between market sensing and market orientation (Source: Foley and Fahy 2004). n. al. er. Text mining is an area of computer science combing data mining, machine. i n U. v. learning, natural language processing and information retrieval, etc. to find out. Ch. engchi. important significant in huge-amount contexts (Feldman and Sanger, 2007). In addition to basic text mining principle about word segmentation, term weighting scheme, vector space model, term document matrix and similarity calculation and document clustering, text mining still has other applications. As a result, text mining is a tool that enables us to apply on distinguishing the importance of data and information on internet relating to our demand for facilitating market sensing capability. The following paragraphs briefly introduce basic concepts of text mining. In order to point out the existing market sensing method in industry, we can simply overview the pros and cons in Table 2-1. 14.

(24) Table 2-1. Comparison of existing methods of market sensing Method. Description. Advantages. Disadvantages. Search. 1. Only useful for Search engine such. Engine. keyword searching as Google and Multiple. 2. The database of. Yahoo is the most information of. important market. multiple databases.. intelligences is not. popular searching method for general open for google people.. Private. 治政Company’s 大 private. 立. search.. Private knowledge Knowledge. information can be. ‧ 國. base. Information is too. 學. base is designed for found, which is not. internal company. limited. ‧. available for other usage. Company. is professional for. n. al. er. Consulting company. io. Consulting. sit. y. Nat. people. Ch. summarizing market intelligences for. engchi. i n U. v. 1. High fee cost. 2. High time cost.. different domains. In other words, our research for enhancing company’s market sensing capability is required to have more efficacy and efficient performance than the three methods of Table 2-1. Promoting market sensing capability in the research is focused to develop operant resource capability, which involves text mining technology in our method, we hence introduce text mining technology in the following paragraph.. 15.

(25) 2.4 Term processing technique In information retrieval view, term processing technique is necessary for analyzing document to extract feature and key word of the document. In the following paragraph, we will explain in two aspects “word segmentation” as well as “term weighting schemes”.. 2.4.1 Word Segmentation:. 政治大 segmentation. N-Gram processes 立 word segmentation in text corpus. If frequency of the. N-Gram text categorization (Cavnar and Trenkle, 1992) is a method of word. ‧ 國. 學. Gram of text corpus is higher than the standard, it would be regarded as independent term. It classifies them into Bi-Gram, Tri-Gram and even N-Gram according to the. ‧. length of neighbor words. For example, there is a sentence “A performance exhibition. sit. y. Nat. of outstanding plan, inviting you to join together”. After performing N-gram algorithm,. al. er. io. the sentence is split by sign “,” to replace non-character. The sentence will be separated. v. n. into “A performance exhibition of outstanding plan” and “inviting you to join. Ch. engchi. i n U. together”. Afterward, performing Bi-Gram to segment the first sentence to twocharacter words character by character and extracting the word combination frequently appearing in the word base, the result words are “Performance”, “Exhibition”, “Outstanding” and “plan” in the final.. 2.4.2 Term weighting schemes. Later on word segmentation, all terms of each document is analyzed. In order to arrange typical term, evaluating the importance of words and calculating weights in different documents is a possible solution. 16.

(26) Term weighting is defined by calculating certain term in local and global weight. Local weight represents importance of a term in certain document, and global term is importance of a term in entirely collective documents. Popesecu (2001) refers formula of local and global weight in Table 2-2 and 2-3. Table 2-2. Local Weight in an article Local Weigh Formula. Name. 1 if 𝐟𝐢𝐣 > 0. 政治大. 0 if 𝐟𝐢𝐣 = 0. 立. Within-document frequency. 學. ‧ 國. Fij. Binary. 1+ log 𝐟𝐢𝐣 if 𝐟𝐢𝐣 > 0. Log. ‧. 0 if 𝐟𝐢𝐣 = 0. y. Nat. (1+log 𝐟𝐢𝐣 ) / (1+log aj) if 𝐟𝐢𝐣 > 0. sit. Normalized log. n. al. er. io. 0 if 𝐟𝐢𝐣 = 0. v i n Ch e n g cAugmented h i U normalized term frequency. 0.5 + 0.5 (𝐟𝐢𝐣 / 𝐱 𝐣 ) if 𝐟𝐢𝐣 > 0 0 if 𝐟𝐢𝐣 = 0. Source: (Preoescu, 2001) Table2-3. Common Formula of Global Weight Global Weight Formula. Name. log (N / 𝐧𝐢 ). Inverse document frequency. log [ (N –𝐧𝐢 ) / 𝐧𝐢 ]. Probabilistic inverse. 17.

(27) Entropy. 𝐅𝐢 / 𝐧𝐢. Global frequency IDF. 1. No global weight. Source: (Preoescu, 2001) In Table 2-2 and Table 2-3, fij represents frequency of term i in document j(Term frequency); aj represents average frequency of the terms in document j; 𝐱 𝐣 represents. 政治大. the most appearing frequency of any term appearing in document j(Document. 立. frequency); N is number of documents in the entire collection set; ni is number of. ‧ 國. 學. documents that term i appearing; Fi is accumulated frequency of term i in the entire. ‧. collection.. Term process technology shows how this research deals with and integrates the. y. Nat. n. al. er. io 2.5 Vector Space Model. sit. texts of market intelligences at the beginning.. Ch. engchi. i n U. v. Vector Space Model (VSM) is the most general information retrieval model. VSM is most simple and powerful with production model (Salton, 1988). VSM is presented by vector space matrix consisting of keywords and documents, and queries and documents are presented by vectors of keywords. However, the result is for calculating the similarity between query and document. By the method of preceding paragraph 2.3.2 “TF-IDF” to calculating term weights, the vector of a document D can be exhibited as D = (w1 , w2 …wn ) (w means the weights of terms in the document ).. 18.

(28) 2.5.1 Term-Document Matrix. Term-Document Matrix is the core of Vector Space Model. Term-Document Matrix can easily express the concept of term and document with machine. In Figure 2-2, a matrix illustrates a representation of k documents with i terms, in which Wik is refers to the weights of term i in document k.. 立. 政治大. ‧. ‧ 國. 學 sit. y. Nat. Figure 2-2. Term-Document Matrix (Salton and Gill, 1983). er. io. 2.5.2 Similarity Calculation. al. n. v i n C h mechanically U Similarity calculation is often applied e n g c h i in the process of clustering and classifying documents. Table 2-4 then shows commonly used similarity calculation formula. Table 2-4. Common Similarity Calculation Similarity Calculation. Evaluation for Binary Term. Evaluation for Binary Term. Formula. Vectors. vectors. Inter product. 19.

(29) Dices coefficient. Cosine Coefficient. Jaccard coefficient. 政治大. 立. ‧. ‧ 國. 學 (Salton, 1988) pp.318. Nat. sit. y. The section vector space model illustrates how this research transforms the. n. al. er. io. document to matrix of market intelligences for data mining procedure. 2.6 Document Clustering:. Ch. engchi. i n U. v. Clustering is a method to classify different documents to different groups with similar features (Lin and Jang, 1998), moreover, finding out their typicality point to reduce data amount. These groups with well-classified are called “cluster”. In principle, there are two types of Clustering method: 1. Hierarchical clustering: The requirement of the clustering method is calculating the similarity between documents in pair is a requirement, while total amounts of documents are incessantly increase, the cost would be prodigious.. However, Salton proposed “single-pass clustering” is one possible solution to calculate little cost. 20.

(30) clustering (Salton and Gill, 1983). The algorithm of Clustering method is: 1. Extracting first document to be the first cluster 2. Extracting next document, then compare their similarity with existing clusters. 3. Allocating the document to appropriated clusters 4. Re-calculating vector of the changed cluster 5. Flagging the document which doesn’t allocate to any existing cluster as a new cluster 6. Repeat step 2 to 6 until finishing all documents. The calculating cost is little so that it is adequate for on-line clustering.. 政治大 the group number before clustering, but the group keeps on augmenting at all hours in 立. 2. Partitional clustering: The requirement of the clustering method is in advance setting. clustering method apparently cost lots of calculating cost.. 學. ‧ 國. dynamic news tracking. This clustering method hence is not suit for our research. The. ‧. This section vector space model depicts the procedure of performing data mining. sit. y. Nat. for the market intelligences.. er. io. 2.7 News Event tracking and detecting. al. n. v i n C horiginally is predominant Topic detection and tracking (TDT) with Defence Advanced engchi U Research Projects Agency (DARPA), which tracks new events from stream of broadcast news like Figure 2-3.. 21.

(31) Figure 2-3.Topic Detecting and Tracking (Panka, 1999). 政治大. TDT program originated from 1996 and presented the final report of pilot study (Allan. 立. et al., 1998). Carnegie Mellon University (CMU) as well as University of. ‧ 國. 學. Massachusetts (UMass) are famous academic research institution on TDT. Moreover, both of them are also responsible for accomplishing TDT pilot study program.. ‧. The event of news is defined an affair happened in certain position and exact time. y. Nat. io. sit. (Allan et al., 1998). For example, “there comes a tornado in Xindian,2011 May twelfth”. n. al. er. is an event, but individually discuss tornado is not counted as an event.. i n U. v. 1. Event detection. It is defined as “Discovering new or unobservable events from. Ch. engchi. successive news stream”. CMU adapted vector space model to present a document and single-pass clustering to form events (Kurt, 2001). Event detection is an unsupervised learning and be classified to two classifications: (1) Retrospective detection: The detection method means digging out unobservable events from given document collection by the same method of CMU detection. Input is a document collection of new affairs, and output is story cluster evaluated. Besides, each story cluster represents an event.“ (2) On-line detection: It stands for successive instant news or earmarking new events of information. Based on CMU’s detection method, input is series of documents in order 22.

(32) by emerging time; output is determination “yes or no” as whether the new document is a beginning of the new event or not. 2. Event Tracking: The goal of event tracking is classifying new documents to previously existing events. News event tracking and detecting shows how this research identifies newest and valuable market intelligences. 2.8 Story, Event and Topic. 政治大 integrated document in database requires transformation. In the flow of IR, forms of 立. In the process of event tracking and detection, raw document from internet to well-. these documents which we extract into different type:. 學. ‧ 國. document information in each stage are regarded as various classifications: We classify. ‧. 1. Story: a paragraph of words to provide user information such as un-integrated news,. sit. y. Nat. blogs, text document.. io. er. 2. Event: a specific affair with emerging date and certain position, which is composed of diverse stories with same characteristics to illustrate a same affair.. al. n. v i n 3. Topic: a collection composed C of related events. A certain h e n g c h i U topic is composed of variety events with the same feature.. The section vector space model illustrates how this research transforms the document to matrix of market intelligences for data mining procedure. 2.9 Market Relating Capability Market relationship is a topic priority for most companies. Market relating capability spans many functions as well as levels in the scope of connecting organization with buyers and sellers. Linking to buyers and sellers is limited to the following visions: 1. Information sharing: opening the lines to communicate and share decisions to 23.

(33) complete Electronic Data interchange. 2.. Close link: while cooperative comates are required, social networks create close links at different levels management. On the other hand, the reality “social encounter is a service encounter” can be capitalized by social links in service market.. 3. Process integration: The generation of toughest connections is happened in the process of two partner designing and managing together. The toughest connections are obtained to get earlier input, share fuller information and integrate more. 政治大 However, scholar Day(1999) proposed a study to examine whether the role in 立. smoothly by the cooperation between firm, supplier and customer.. market relating capability influences the development of relationship-based. ‧ 國. 學. competitive strategy. Day addressed three elements of market relating capability as. ‧. Figure 2-4. The three elements “Orientation”, “Integration and alignment of process”. sit. y. Nat. and “Knowledge and skill” enhance and interact with each other, in addition, they can. io. Orien ta tion. n. al. Ch. engchi. er. separately be handled with evaluating organization capability and change program.. i n U. v. C olla bora tive P a rtn erin g C a pa bility. Tra n s a ction a l C a pa bility. Va lu e-a dded R ela tion s h ip C a pa bility. K n owledge a n d S kills. In tegra tion a n d a lign men t of proces s es. Figure 2-4. Three elements of market relating capability(Day, 1999) In this model, if firms move each vectors further, they will achieve the capabilities 24.

(34) required for their cooperating partners more. The three elements are then briefly introduced below: 1. Orientation: it comprises mind-set, value and norms of organization which affect customer interaction between customers. 2. Integration and alignment of process: it is the concept about integrating and aligning interaction technology in activity process. 3. Knowledge and skill: it represents the fundamental knowledge related to customer relationship and professional skill.. 政治大 organizational alignment and making learning are a priority to go on improving; 立. After Engaging both minds and hearts of the entire organization, tightening. organization hence gets benefits such as more successful relationship-building. ‧ 國. 學. initiative, better risk of imitation, better defense by continuously learning and. ‧. innovating and more difficult obstacle for rivals to grasp the organizational secret.. sit. y. Nat. For enforcing market relating capability, chance discovery is a possible method to. io. er. jointly design or discover problem among users from different groups in activity process. In the process of chance discovery, keyGraph algorithm assists people in. al. n. v i n C h inside and outside discovering the relationship among organization’s data (business engchi U. situation and market tendency) and new chance from people’s mind (business goal). The further discussion of chance discovery theory will be illustrated in the next section. On the other hand, to point out the existing market relating method in industry, we can simply overview the pros and cons in Table 2-5. Table 2-5. Comparison of existing methods of market relating Method. Description. Advantages. 25. Disadvantages.

(35) 1. Additional human Gathering all. Deeply discussing. participators and. the issues of. engaging the. business. resource cost to record or transform Face-to-face. the documents. brainstorming. discussions. development,. together with a. market trend and. physical method.. ability of company.. 2. Hard to find the relationships among different information. 1. Automatically recording all 政治大 system or tool to the documents. 立. Using information. tool for. Hard to find the relationship among. 2. Discussing. participators with a. with more. virtual method.. participators. 學. Brainstorming. discuss with all. ‧ 國. Information. different information.. ‧ sit. y. Nat. online.. io. al. er. That is, our research dedicated in enhancing company’s market relating capability. n. must have more efficacy and efficient performance than both the methods of Table 25.. Ch. engchi. i n U. v. Promoting market relating capability in the research is another emphasis to develop operant resource capability, which involves chance discovery theory in our method; we hence introduce chance discovery theory in the following paragraph. 2.10 Chance Discovery Chance Discovery is a study of discovering a chance as event significant for decision making (Ohsawa, 2003), which helps user find the hidden exclude in the data by data mining. In scenario map view, we can regard collections of co-occurrent events as an island. Bridge plays the important role of connecting different islands, which usually 26.

(36) comprise of scenario. The goal of Chance Discovery is questing significant bridge include/hidden in inside/outside data. The primary methods to perform Chance Discovery are mentioned below:. 2.10.1. KeyGraph. It is a method to visualize structure of words as a map through KeyGraph algorithm to depict items as nodes in the data (Ohsawa, 2005). In other literatures, keyGrapgh was also utilized to analyze competitive relationship between companies (Hong, 2009),. 政治大 Ohsawa & Okazaki, 2006),立invisible relevant persons in. development opportunity from the relationship of company’s product patent (Horie, a homogeneous social. ‧ 國. 學. network (Maeno, Kiichi Ito & Yukio Ohsawa, 2007) and the innovative opportunity from the relationship among creative products (Wang, Ohsawa & Nishihara, 2012). For. ‧. instance, to form texts of six meetings into a data as an island(see Figure 2-5),. sit. y. Nat. KeyGraph method visualizes the event map in Figure 2-5. Black nodes represent most. n. al. er. io. frequent items in the data, and red node represents rare hubs bridging item (e.g.. i n U. v. member4 and member9) to connect island{member1, member2, member3} and island{. Ch. engchi. member5, member6, member7} as co-occurrence collection.. Figure 2-5. Data of a cluster. 27.

(37) Member 3 Member 4. Member 2. Member 5. Member 1 Member 9. Member 7. Member 6. 治政大 Method Figure 2-6. an Example of KeyGraph 立. After adding inside data from user to additionally generate new event map,. ‧ 國. 學. constantly evaluating, discuss results most valuable bridge nodes with scenario.. ‧. 2.10.2. Data Crystallization. sit. y. Nat. n. al. er. io. It presents hidden chance not included in the data structure. The two primary methods. i n U. v. are inserting dummy items as unobservable or totally unknown events, and human’s. Ch. engchi. process of discovery where chance might not in the data. A tool Data Crystallizer iteratively inserts artificial dummy nodes to present potential unobservable event to the given data and revises in previous case. The size of each data enlarges due to reducing the granularity of structure in each iteration. After this procedure, it obtains new data set with structure same as hidden event connecting to each level. In human-machine interaction, after particular users have deeply understood the complex product by data crystallization, the concept what users express will be visualized to the scenario map. All users start to evaluate and improve those dummy nodes. Repeating these procedures,. 28.

(38) if all the users agree with the dummy nodes, they might be new chances in the event map. This section illustrates how this research promotes market relating by finding the relationship among words and inserting dummy nodes. Through the chance discovery theory, SMEs can speed up their market relating capability to generate new chances for their insight.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 29. i n U. v.

(39) CHAPTER 3 PROSPERITY TAIWAN INTEGRATED PROJECT. The research work of the thesis is actually a part of a research project dubbed “Prosperity Taiwan”. The primary goal of “Prosperity Taiwan” is to create a set of ITenabled policies and workflows that facilitate Taiwan’s economic transformation. Under this project, the research dedicates to create an IT artifact “Sourcing+”. Sourcing+ plays an important role in searching related market intelligence and coming up with market-oriented chance to satisfy people’s motivation and make new strategy.. 政治大 3.1 Taiwan’s Economic Strengths and Current Economic Background 立. ‧ 國. 學. Taiwan, as one of the Four Asian Tigers, which economic growth from 1960s to 1990s. ‧. has been exceptional, maintaining at a growth rate excess of 7% per year. Taiwan’s economic strength stems from its manufacturing capabilities, especially in terms of. y. Nat. io. sit. information technology related products. Bulk of Taiwan’s business comes from. n. al. er. building and manufacturing white-label products that is later branded with American. Ch. i n U. v. or European brands. However, Taiwan’s greatest economic strength turns out to be its. engchi. biggest liability in recent years due to changes in the global economic landscape. The current global economic landscape presents several challenges for the Taiwanese economy: increase in cheaper manufacturing alternatives such as manufacturers from China or Brazil, leading to increased price competition and severely cutting profit margins for Taiwanese manufacturers. There is also an increased premium placed on brand recognition and intellectual property; despite Taiwanese’s manufacturing prowess when it comes to building world class hardware such as iPhone, iPods and other globally recognized devices, they often merely receive up to 3 to 5% profit share of each device sold. The reason is that the Taiwanese manufacturers do not 30.

(40) own the brand and or intellectual property of the device. For responding the above challenges, the Taiwanese government emphasized several industries poised for growth including medical services, bio-technology, green energy, culture/arts/creativity and high technology agricultural. 3.2 The “Prosperity” in Prosperity Taiwan Prosperity comes in various forms. While most people associate “Prosperity” with material wealth, we are interested in creating all-round prosperity. They are namely:. 政治大. Material Prosperity, Spiritual Prosperity, Physical Prosperity and Social Prosperity.. 立. 3.3 Vision of Prosperity Taiwan. ‧ 國. 學. The goal for Prosperity Taiwan is to create sustainability in economic development by. ‧. valuing up the value chain. The five major goals are as follows: Enhance university-industrial cooperation. 2.. Kick start Taiwan’s economic transformation by creating “industrial modules”. 3.. Create a model for industry transformation. 4.. Start small, go big. 5.. Internationalize based on successful industrial transformation. n. al. Ch. engchi. er. io. sit. y. Nat. 1.. i n U. v. 3.4 Culture, Arts and Creativity as an Example At the outset, Prosperity Taiwan focuses on the culture, arts and creativity industry. The research team believes that the industries of culture, arts and creativity are Taiwan’s advantages due to Taiwan’s unique legacy and hybrid of Chinese and Western influences. We also believe that these groups and their related stakeholders will benefit most from our research work. We find that the following characteristics define the culture, arts and creativity 31.

(41) industry: . No limited liability or private limited companies. . They have limited ability to scale its production capabilities. . Typically lack financial support or lack knowledge on funding options. . Business owners are at best conservative regarding the growth prospects of the culture and arts On an industry level, the market exhibits the following characteristics:. . Market lacks a system for evaluating intangible assets. . Industry lacks interdisciplinary talent. . 學. . The industry lacks economies of scale to support growth. ‧ 國. . 政治大 Lack of financial support or lack transparency in terms of funding options 立 Multiple stakeholders with political, economical, social and environmental. ‧. objectives.. sit. y. Nat. To struggle for the above issues, we propose the use of intelligent service machines. io. n. al. er. to aid industry transformation.. i n U. v. 3.5 Intelligent Service Machines to aid Economic Transformation. Ch. engchi. Intelligent Service Machine(s) (Tung and Yuan, 2010) refers to an intelligent design of the service machine featuring the embodied cognition of co-production in terms of modeling and automating the cognitive process and knowledge representations as required. Conceptually, a simple service machine is a socio-technical system (STS) that involves people, model, architecture, technologies for modeling and automating service processes.. 32.

(42) Figure 3-1. A simple service machine is a hybrid of STS and SSME 3.5.1 The V+ Platform. 立. 政治大. ‧ 國. 學. The V+ platform (Figure 3-2) is an intelligent service machine that comprises of 5 simple service engines assist industry transformation in solving problems as. ‧. exemplified in the independent movie makers and the industry as a whole.. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. Figure 3-2. The V+ Platform From a macro-level view, we have Capital+ forming the base of the V+ platform because every stage of a product life cycle in various industries will most likely involve some form of funding activities. On top of Capital+, next, we have Discover+ which helps businesses discover new insights, which can be further strengthened or extended by the design insight patterns recommended by Design Intelligence+. Finally, Sourcing+ can readily translate design insights into market opportunities by bringing 33.

(43) about various operant resources from possible stakeholders in an industry under a holistic system thinking view. Finally, Intellectual Property (IP+) comes into play by ensuring that proper IP strategies can be formed based on the operant resources identified in Sourcing+. From micro-level view, each engine works as such: 1.. Capital+: designed to solve the problem of funding and investment on an individual level and the industry as a whole. It provides a prediction model for individuals to estimate which investor(s) has a higher possibility of funding them.. 政治大 evolution of investors and companies within an industry. 立. In addition, Capital+ provides general industry growth models based on the. Discover+: an IT-artifact with common sense knowledge aimed to facilitate design. 學. ‧ 國. 2.. insight discovery process. The common sense knowledge in this engine is powered. y. sit. Design Intelligence+: is developing the framework and method to externalize,. io. er. 3.. Nat. with only the goods-dominant logic mindsets.. ‧. by ConceptNet. Discover+ is targeted at assisting novice and enterprise designers. formalize and facilitate “insight combination” that is a process of discovering. al. n. v i n C h and “design patterns” “insights” and engchi U. connection between. combining them to. generate idea or inspiration for service design. 4.. Sourcing+: a service machine targeted at Small and Medium Sized enterprises (SMEs) to enable them to discover niche markets when given design insights. Sourcing+ is also an IT-enabled system that exploits text mining technology and chance discovery theory to search for open new or appropriate operant sources from various stakeholders to provide discovery of market opportunities found in different market.. 5.. Intellectual Property: aiming at fostering the use of intellectual property as a means for moving up the value chain either through forging alliances or as a 34.

(44) litigation/patent strategy according to the identified operant resources by Sourcing+. It provides a framework based on value network that help companies to pursue, manage and maneuver their IP strategy. V+ platform is designed to act as a generic service engine that can be applied to most situations. For simplicity and demonstration sake, we explain an application scenario of how V+ platform can be applied in the next section.. 3.5.2 Application Scenario of V+ platform. 政治大 social factors when a company 立 or business sets up a factory in a less developed region. Consider the various issues ranging from political, economical, environmental and. ‧ 國. 學. Very often, the company may face protests from local residents or even local governments due to concerns over environmental pollution, negative impact on way of. ‧. life of local residents or even adverse effect on the local economy. The V+ platform. sit. y. Nat. works by helping various stakeholders search and suggest possible solutions to this. n. al. general public on environmental protection.. Ch. engchi. er. io. problem. Assumption of the agreed solution is to create a museum to educate the. i n U. v. In this scenario, we are expecting multiple stakeholders and not all funding for building the museum will come from the company. Therefore, V+ platform can help to source for multiple investors based on the stakeholders involved. This can happen at various stages of the construction of the museum. The main engine used for sourcing investment is done via Capital+. Discover+ creates new insights based on the stakeholders’ objectives. These stakeholders include the local residents, governments, potential visitors to the museum and the company involved. After gathering the stakeholders’ requirements, these requirements are used as inputs for the Disvoer+ engine, where new and potential. 35.

(45) insights maybe created for framing and reframing process. Such information is useful for museum designers or related service designers. The output of Discover+ is then used as an input for Sourcing+. Design Intelligence+ then helps elevate the insights to foster the degree of impacts in terms of the scalability of stakeholders involved or the service values incurred. Sourcing+ analyzes the insights provided by Discover+ and maps out possible required market intelligence and or related trends. For example, what can one expect from a museum? What form of services, artifacts, displays, etc. can one expect from a. 政治大 integrating current trends and expectations, such as screening documentaries or 立. museum given various stakeholders? Sourcing+ helps answer these questions by. educational movies on environmental protection, interactive learning for children and. ‧ 國. 學. so on as inputs for market sensing. The market sensing capabilities of Sourcing+ then. ‧. facilitates research between various stakeholders and generates possible answers to. sit. y. Nat. aforementioned questions.. io. er. Intellectual Property+ adds support to our scenario by providing IT-enabled support for understanding the overall patent value based on the operant resources. n. al. Ch. identified by Sourcing+. Take for instance, the. engchi. v i n museum U might. be interested in. deploying an operant resource of interactive TV. The related technologies and hence patents may be related to such as those used in Microsoft’s Kinect. Such knowledge will help potential implementation avoid litigation and hence pave the overall patent/litigation strategy.. 3.5.3 Application Scenario of Sourcing+. James is a designer in company A and the current mission of his department was to make a solution to help children as their customers to inspire more motivation to receive. 36.

(46) knowledge of environment production. Within V+, Capital+ can help James find possible sponsors for his current condition, which would give him more confidence to establish his work. Discover+ can aide James in generating significant insight for his solution (as Figure 3-3). After getting the insight as an input from Discover+, Sourcing+ (i.e., this research's system) can assist James with searching related market trend or market intelligence in dynamic market.. 政治大. 立. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. i n U. v. Figure 3-3. The Service Process of Value+ Engine. Ch. engchi. In Sourcing+, the first important capability of market sensing capability would analyze the insight into different trendy topics in terms of text mining pertinent resource information on internet. As we can see in Figure 3-4, the market trends might include: 1. Government environment resource development center 2. Scholars who advocate environment protection 3. Mini movie for learning behavior 4. Interactive learning for children 5. Learning website for children, and so on. 37.

(47) 立. 政治大. Figure 3-4. Different Type of Market Intelligence. ‧ 國. 學. Sourcing+ integrates these stories into events as market intelligence for user’s value creation, for instance, related stories of “interactive learning for children” such as. ‧. interactive screen wall, Smart film compose an event “an interactive technology. y. Nat. io. sit. company Maxiken with Customization function” (Figure 3-5), the event is about a. n. al. er. company with famous interactive technology and low-priced customization capability in Taiwan.. Ch. engchi. i n U. Figure 3-5. Instances of event 38. v.