200945224 - 九、發明說明： 【發明所屬之技術領域】 本發明係涉及一種運用模糊理論之企業資源規劃績效 評估系統及其績效評估方法，特別是指一種利用模糊理論 和群體決策方法之創新設計者。 【先前技術】 按，企業資訊系統(Enterprise Information System ， ◎ EIS )如：企業資源規劃（Enterprise Resource Planning， ERP )、供應鏈管理（Supply Chain Management，SCM )等 已於現今企業體中廣泛使用，藉以提昇企業運作效率、增 進企業優勢。 導入所述企業資源規劃系統往往需要投注龐大人力、 財力與時間。而且，所述企業資源規劃系統的運作績效， 亦將深遠地影響企業的表現。過去至今，已經有許多的研 _ 究針對企業資源規劃系統的績效指標選用、使用者滿意度 、資料的品質等進行研究；昔之資訊系統之績效研究，主 要鎖定於績效指標之分類與選用（Delone and McLean，1992 ;Myers et al. ； 1997 ； Skok et al. > 2001 ； Heo and200945224 - Nine, invention description: [Technical field of invention] The present invention relates to an enterprise resource planning performance evaluation system using fuzzy theory and its performance evaluation method, in particular to an innovative designer using fuzzy theory and group decision making method . [Prior Art] According to the Enterprise Information System (EIS), such as Enterprise Resource Planning (ERP) and Supply Chain Management (SCM), it has been widely used in today's enterprises. In order to improve the efficiency of business operations and enhance the advantages of enterprises. Importing the enterprise resource planning system often requires a lot of manpower, financial resources and time. Moreover, the operational performance of the enterprise resource planning system will also profoundly affect the performance of the enterprise. In the past, many researches have been carried out on the selection of performance indicators, user satisfaction, and quality of data in enterprise resource planning systems. The performance research of information systems in the past has been mainly targeted at the classification and selection of performance indicators ( Delone and McLean, 1992; Myers et al.; 1997; Skok et al. >2001; Heo and
Han, 2003)、財務績效分析（Hunton et al.， 2003)、使用 者滿意度分析（Klenke，1 992 ; Saarinen， 1996 ; McHan ey et al· ， 2002)、資料與系統之品質（Lee et al· ’ 20 02 ； Xu et al. ， 2002)等；然而，仍舊少有研究或實務 案例能夠提出一套具體可行、切合需求的企業資源規劃系 5 200945224 - 統績效評估方法。 _ 是以，針對上述需求，如何研發出一種確具理想實 性之企業資源規劃績效評估系統及其績效評估方法，實 待相關業界再加以思索突破之目標及方向者。 有鑑於此，發明人本於多年從事相關產品之製造開 與設計經驗，針對上述之目標，詳加設計與審慎評估後 終得一確具實用性之本發明。 ❹ 【發明内容】 即，本發明之主要目的，係在提供一種運用模糊理 之企業資源規劃績效評估系統及其績效評估方法，其所 解決之問題點，係針對如何研發出一種具體可行、切合 求的企業資源規劃系統績效評估方法為目標加以思索突 者； 本發明解決問題之技術特點，主要在於提供一種企 資源規劃績效評估系統模糊式績效評估方法，其大致包 Ο 有：績效評估指標之選擇與架構階段，係組成績效評估 容發展小組以負責績效指標之選定與發展出一致的評估 準，並擴展企業資源規劃績效評估系統之導入目標成為 績效評估指標，復進行績效指標數目之增減調整以建構 績效指標構面；群體模糊決策階段，係經由績效指標評 内容與方法後，進行模糊群體績效評估，並採用定量或 意方式評估，所述語意評估值係轉換成對應之模糊數。 估者將所收集資料以語意值集合進行績效指標之語意評 後，復進行語意評估值彙總，以求得每一主要構面綜合 用 有 發 論 欲 需 破 業 括 内 標 其 出 估 語 評 估 模 6 200945224 Ο 糊評估值和整個系 與系統改進階段， 系統改善方案，其 積分數法將構面的 糊評估值轉換回語 藉以將轉換的語意 體糸統績效之變化 效評估方法得兼顧 且採用集體評估決 觀意識，進而能獲 統績效的總模糊評估值； 其步驟包括評估成果分析 中該評估成果分析討論階 綜合模糊評估值和整個系 意值’·而所述擬定系統改 值結果依據時間的推移產 趨勢；藉此創新設計，使 定量與定性指標，使績效 策方式給定評估值，減少 得更加客觀準確而實用的 評估結果分折 討論以及擬定 段可利用模糊 統績效的總模 善方案階段係 生各構面..和楚 本發明所揭績 指標完整，並 評估者個人主 評估結果。 【實施方式】 請參第1圖所示，係本發明運用模糊理論之企業資源 規劃績效評估系統及其績效評估方法之軔 ’、“、 — 心权佳實施例，惟此 荨貫施例僅供說明之用’在專利申請上 此+焚此結構之限 制；本發明的主旨係為發展一個企業資 貝碌規劃績效評估系 ⑩ 統（以下簡稱ERP )之績效評估方法，太么从 轚 本糸統開發完成後 之應用面，將可安裝於企業之管理資訊邱 札邛門，配合ERP系 統之運作’作為週期性ERP系統績效評估 τ估之用’所輸出之 資訊’將可提供給管理者監控ERP系絲达& $、、死％致之變化與對 ERP系統進行改善方針之擬定。 '、 其方法部份如第 坪估指標（Perfo (二）群體模糊 進階段。兹依據 整體系統運用於ERP系統績效評估， 1圖所示，係包括三個階段：（一）績效 rmance Indicators; Pis )之選擇與架構階段、 決策階段、（三）評估結果分析與系統改 7 200945224 - 此流程將系統開發之理論與步驟流程說明如下。（整體系 _ 統流程請配合第2圖之簡易文字方塊所示） (一）績效評估指標之選擇與架構階段，其步驟包括： 1. 組成ERP績效評估内容發展小組：ERP績效評估内容 發展小組（ERP PI content panel)應包括重要的管理者、使 用者代表、系統的專家和顧問等，其主要負責績效指 標（Pis)之選定與發展一致的評估標準。 2. 擴展ERP導入目標成為績效評估指標：公司在ERP系 〇 統選擇與導入階段時定義企業導入ERP之目標，這些 目標指出了管理者希望導入ERP系統後所達成的事項 。因此ERP系統績效評估指標必須能夠反應這些所設 . 定目標的達成度。RP績效評估内容發展小組由所定義 好的ERP系統導入目標出發，討論其適合的績效評估 指標。如此，討論完每個目標對應的所有績效指標， 即建立起最初的績效評估指標集合（Original PI set)。 3. 由產出面增加適當的績效指標：由於初步的績效評估 Q 指標集合中的績效評估指標來自於ERP導入目標架構 的延伸，屬於投入面的考量，可能並不足夠用來衡量 ERP系統的產出面。過去對於資訊系統之績效評估的 研究中，許多學者提出許多的績效指標，這些績效指 標大部分均在考量系統之產出面之衡量。因此建議參 考過去的文獻所使用的績效評估指標加以補充，彌補 由輸入面的ERP導入目標架構轉化的績效指標的不足 〇 4. 刪減績效指標之數目：由於太多的績效指標將導致評 8Han, 2003), financial performance analysis (Hunton et al., 2003), user satisfaction analysis (Klenke, 992; Saarinen, 1996; McHan ey et al., 2002), data and system quality (Lee et al · '20 02 ; Xu et al., 2002); however, there are still few research or practical cases that can propose a set of concrete and feasible and appropriate enterprise resource planning systems. _ Therefore, in response to the above requirements, how to develop an enterprise resource planning performance evaluation system and its performance evaluation method that are ideally realistic, and wait for the relevant industry to reflect on the goals and directions of breakthrough. In view of this, the inventor has been engaged in the manufacturing and design experience of related products for many years, and has made the practical and practical invention after detailed design and careful evaluation for the above objectives. ❹ 【Abstract】 The main purpose of the present invention is to provide an enterprise resource planning performance evaluation system and its performance evaluation method using fuzzy theory, and the problem solved by the present invention is to develop a concrete feasible and suitable method. The enterprise resource planning system performance evaluation method is aimed at thinking about the target; the technical characteristics of the problem solving of the invention is mainly to provide a fuzzy performance evaluation method for the enterprise resource planning performance evaluation system, which generally includes: performance evaluation index In the selection and architecture phase, the performance evaluation team is responsible for selecting and developing the performance indicators to achieve a consistent evaluation criteria, and expanding the introduction goal of the enterprise resource planning performance evaluation system to become a performance evaluation indicator, and increasing the number of performance indicators. Adjustment to construct the performance indicator facet; in the group fuzzy decision-making stage, after evaluating the content and method through the performance indicators, the fuzzy group performance evaluation is carried out, and the quantitative or intentional evaluation is adopted, and the semantic evaluation value is converted into the corresponding fuzzy number. The estimator collects the collected data from the semantic value of the performance indicators, and then summarizes the semantic evaluation values to obtain the evaluation of each major facet.模 6 200945224 Ο paste evaluation value and the whole system and system improvement stage, the system improvement program, the integral number method converts the facet evaluation value into a sentence to take into account the conversion effect of the semantic performance Collectively assessing the awareness of the decision, and then obtaining the total fuzzy evaluation value of the overall performance; the steps include analyzing the evaluation result in the evaluation result analysis and discussing the comprehensive fuzzy evaluation value and the whole system meaning value. According to the passage of time, the production trend; with this innovative design, the quantitative and qualitative indicators, the performance evaluation method to give the evaluation value, reduce the more objective and accurate and practical evaluation results, and the total number of available fuzzy performance The mode of the good project is the various facets of the students. Estimated results. [Embodiment] Please refer to Fig. 1 for the example of the enterprise resource planning performance evaluation system and its performance evaluation method using fuzzy theory in this invention, and the best example of the heart right. For the purpose of explanation, 'there is a limitation on this patent in the patent application; the main purpose of the invention is to develop a performance evaluation method for the enterprise performance evaluation system (hereinafter referred to as ERP), too After the development of SiS, the application surface will be installed in the management information of the company. Qiu Zhamen, with the operation of the ERP system as the information of the periodic ERP system performance evaluation τ estimation will be available for management Monitor the ERP system wire & $, change the death rate and the improvement of the ERP system. ', the method part of the ping evaluation index (Perfo (b) group fuzzy stage. Based on the whole The system is applied to ERP system performance evaluation. As shown in Figure 1, it consists of three phases: (1) performance rmance indicators; Pis) selection and architecture stage, decision stage, and (3) evaluation results. Analysis and System Reform 7 200945224 - This process describes the theory and process flow of system development as follows. (The overall system is shown in the simple text box of Figure 2) (1) The selection and structure stage of performance evaluation indicators, The steps include: 1. Forming an ERP Performance Assessment Content Development Team: The ERP PI content panel should include key managers, user representatives, system experts, and consultants, who are primarily responsible for performance indicators ( Pis) is selected and developed in accordance with the evaluation criteria. 2. Expanded ERP import target becomes performance evaluation index: The company defines the goal of enterprise ERP import in the ERP system selection and import stage. These targets indicate that the manager wants to import ERP system. The matters that are achieved afterwards. Therefore, the ERP system performance evaluation indicators must be able to reflect the achievement of these set goals. The RP performance evaluation content development team starts from the defined ERP system import target and discusses its appropriate performance evaluation indicators. In this way, after discussing all the performance indicators corresponding to each target, the initial performance is established. Evaluation of the indicator set (Original PI set) 3. Appropriate performance indicators from the output side: As the preliminary performance evaluation Q indicator set in the set of performance evaluation indicators from the ERP import target structure extension, is an input consideration, may It is not enough to measure the output of the ERP system. In the past, in the research on the performance evaluation of information systems, many scholars proposed many performance indicators, most of which are based on the measurement of the output of the system. Refer to the performance appraisal indicators used in the past literature to supplement the lack of performance indicators transformed by the ERP import target structure on the input side. 4. Decrease the number of performance indicators: too many performance indicators will lead to the evaluation 8
200945224 估過程的複雜和欠缺效率，因此必須針對績效指 合中的績效指標數目進行刪減。為簡化刪減過程 得群體之共識，將利用 quantitative content validity d (Lawshe, 1975)進行績效指標數目删減。ERP績效 内容發展小組成員對於每一個績效指標進行主觀 等給分：「不重要」給1分、「重要但不是必要 2分、「必要的」給3分。然後利用以下公式計 一個績效指標的 content validity ratio (CVR): CVR = {n-N I2)I{N !2) 其中，岸代表所有的成員數。；7為小組成員給該 指標為2或3分的人數。所有的績效指標均求得 VR，以顯著水準α = 〇.〇5下，求得CVR之門檻值。 門檻值比較後，CVR小於此門植值的績效指標將 除。 5.建構適當的績效指標構面：將剩下的績效指標進 織的工作’以利有糸統地進行績效分析。為了與 系統評選階段的ERP導入目標架構相互一致，所 績效指標構面區分成以下四大範嘴： (1)專案管理之績效指標：牽涉ERP導入專案之管 效的績效指標，例如：專案成本控制、導入時 制等。 (2)系統方面之績效指標：代表ERP系統運作之績 例如：糸統功能、糸統彈性、可靠产、使用者 等。 (3)供應商方面之績效指標：說明ERP系統供應商 標集 與取 met ho 評估 的3 」給 算每 績效 其C 與此 被刪 行組 ERP 以將 理績 程控 效， 介面 之配 9 200945224 - 合績效，如：供應商技術能力、供應商的服務 - (4) 產出影響方面之績效指標··討論ERP系統對組 個人和客戶之影響的績效指標。 所述績效指標階層架構可包含有四個階層：最上 終極目標，即整體ERP系統績效評估。第二層為 主要構面，即專案管理、系統、供應商與產出影 四大主要構面。由每一個主要構面所建立的主要 形成第三層架構。第四層即為欲達成各目標所對 φ 績效評估之績效指標。 其中，對於ERP系統之績效評估，其評估指標可 分成定量與定性兩大類的績效指標。評估者對於 指標的評估可以從該指標之量化數據直接得到其 :但是，因為各定量指標的單位不同所以無法直 算，因此可連行討論訂出各績效指標的評估值的 轉換方式，藉由評估值的落點來轉換成為分數， 數沒有單位，所以可以進行計算與彙總。而定性 ❹ 效指標無法或較難以一個明確的數值量化表示， 者可採用語意（linguistic)方式進行評估；評估小 員依據評估指標架構、定義與評估内容進行資料 集與解讀，給予語意值之討論與評估。 (二）群體模糊決策階段，其步驟有： 1. 發展詳細的績效指標評估内容與方法 在績效指標構面中，一個績效指標是一個可以衡 項目，其代表的是一個目標的達成程度。但是為 等。 織、 層為 四大 響等 目標 應的 以區 定量 數值 接計 分數 此分 的績 評估 組成 之收 量的 了避 10 200945224 不同的評估者有不同的認知，而產生評估的偏差， 所以V f '、須針對每一個績效指標的資料收集方法、評估 早位 i 、評估方法與標準進行討論。在此步驟中，其先 刖在珠 、、 雙立ERP導入目標架構時所累積的知識，有助於 績效：tt , 表久Θ標評估方法和標準的建立。一個標準的問卷或 可以被發展出來協助評估者進行評估。在此同時 —— 。以討論與決定各績效指標的權重。績效指標的權 可由小組成員主觀給定，或藉由兩兩指標相互比較 ❿ 2. ’接、疋，例如利用 ahp (anaiytic hierarchy process) 方法（Saaty ’ 1980)計算指標相對權重。 進行模糊群體績效評估 由於一些的績效指標無法獲得一個明確的數值量化表 示’所以將採用語意（linguis1;ic)方式進行評估。所描 述的語意評估值將轉換成對應的模糊數（fuzzy number) 依據模糊數的性質和擴張原理（Iq ir and Yuan . 1995) ，假s又兩個正規二角模糊數（N〇rmai triangUiar fuzzy numbe 〇 r) g = (〜0'，c】）和5 = ，則其模糊代數運算如下： (1) 右㊉尹2 =(a,+a2，h+62,4+c2) (2) 7；®f2 =(α, χα2,Ζ>, xb2,c] xc2) 評估者將所收集之資料以叉={^P，P，^G，FG}之語意值集合 進行績效指標之語意及權重值評估，其中的績效語意值分 別為 VP =很差（Very Pool)、L =差（Pool) 、F =普通（Fair) 、G =好（Good) 、VG =很好（Very Good)。其所對應之正規三 角模糊數如以下表所示。而其語意權重值也可由小組共同 11 200945224 -決定，其中的績效語意值分別為VL =很差（Very Low)、L = •差（）M k 通（Median)、Η =高（High) 、VH =很高（Very Η_) ’而語意權重值可包括VL =很差（v町l〇w)、l =差 (Low)、Μ =普通（Median)、H =高（High) 、VH =很高（Very 耵 gh);所對應之三角模糊數如下表i所示： 語意權重值和語意評估值轉三角模糊數表 ❹ 語意權重值 __ ·. 很低（VL) 低 （L) 中等（M) 高 （H) 很高（VH) 語意評估值 很差（VP) 差 （P) 普通（F) 好 (G) 很好 (VG) 模糊數 (0，0，2.5) (0, 2.5, 5.0) (2.5, 5.0, 7.5) (5.0, 7.5, 10.0) (7.5, 10.0,10.0) 另者，若評估者對於此語意值轉換對應的模糊數有 不同於表1之區間，本系統亦可由評估者自行設定。再 補充說明的是’模糊理論（fuzzy set theory)係利用「隸屬 函數」（membership function )值來描述一個概念的特質， 亦即使用0與1之間的數值來表示一個元素屬於某一概 念的程度，這個值稱為該元素對集合的隸屬度，而常用 的模糊數型態包括L-R模糊數（L-R fuzzy number )、梯形 模糊數（trapezoidal fuzzy mimber)與三角模糊數（triangular fuzzy number)。為使計算簡便，本發明將以Lee and Park (19 97)所提出的 Efficient fuzzy weighted average algorithm 進行 語意評估值彙總，以求得每一主要構面綜合模糊評估值 (Aggregated fuzzy rating)和整個系統績效的總模糊評估值 12 200945224 • (Total fuzzy rating)。 (二）评估結果分析與系統改進階段，其步驟有： 1. 評估成果分析討論 首先本發明將利用Liou and Wang (1992)所提出的模糊積 分數法（Fuzzy total integral value method)將每一主要構面 的综合模糊評估值和整個系統績效的總模糊評估值轉 換回語意值。利用此模糊積分數法之原因為and Wre (2001a ， 2001b)將至2〇〇〇年所發表的四個重要的 杈糊排序法，並利用9個主要的數學基本定理加以驗 證，發現模糊積分數排序法符合8項定理，為所有模 糊排序方法中最佳者。 假設d為一個模糊數，以其隸屬函數之峰值1分成左 右兩部分’則左隸屬函數定義為片和右隸屬函數定義 為八，且# —[(U] , // ~>[〇，1]。定義4為片之反函數 且&為//之反函數。則2之左積分值（left integral val Ο ue)定義為： 乂之右積分值（right integral value)定義為： 再利用樂觀參數β整合左右二積分值得到2的全積 13 200945224 分值（total integral value)200945224 The complexity and lack of efficiency of the assessment process must be reduced for the number of performance indicators in the performance indicator. In order to simplify the consensus of the group, the quantitative content validity d (Lawshe, 1975) will be used to reduce the number of performance indicators. ERP Performance The content development team members give a subjective rating for each performance indicator: “not important” gives 1 point, “important but not necessary 2 points, and “necessary” gives 3 points. Then use the following formula to calculate the content validity ratio (CVR) of a performance indicator: CVR = {n-N I2)I{N !2) where shore represents the total number of members. ;7 is the number of members of the group who give the indicator 2 or 3 points. All performance indicators are VR, with a significant level of α = 〇. 〇 5, to find the CVR threshold. After the threshold is compared, the performance indicator with a CVR less than this threshold will be removed. 5. Construct an appropriate performance indicator facet: Engage the remaining performance indicators into a “performance analysis”. In order to be consistent with the ERP import target architecture in the system selection stage, the performance indicator facets are divided into the following four categories: (1) Project management performance indicators: performance indicators related to the effectiveness of the ERP import project, such as: project cost Control, import time system, etc. (2) Performance indicators for the system: represent the performance of the ERP system, such as: system function, flexibility, reliable production, users, etc. (3) Performance indicators of suppliers: Describe the ERP system supply trademark set and the evaluation of the 3" for each performance C and the deleted group ERP to control the performance process, interface matching 9 200945224 - Performance, such as: supplier technical capabilities, supplier services - (4) performance indicators in terms of output impacts. · Performance indicators that discuss the impact of the ERP system on individual groups and customers. The performance indicator hierarchy can include four levels: the top ultimate goal, which is the overall ERP system performance evaluation. The second layer is the main facet, namely project management, system, supplier and output. The third layer architecture is formed primarily by each major facet. The fourth level is the performance indicator for the performance evaluation of φ to achieve each goal. Among them, for the performance evaluation of ERP system, its evaluation indicators can be divided into quantitative and qualitative performance indicators. The evaluator's assessment of the indicator can be directly obtained from the quantitative data of the indicator: However, because the units of each quantitative indicator are different, it is not straightforward, so it is possible to discuss the conversion method of the evaluation value of each performance indicator by means of evaluation. The falling point of the value is converted into a score, and the number has no unit, so calculation and summary can be performed. However, the qualitative and effective indicators cannot be or are difficult to quantify with a clear numerical value. The linguistic method can be used for evaluation; the evaluation staff can conduct data collection and interpretation based on the evaluation index structure, definition and evaluation content, and give a discussion of semantic values. And evaluation. (2) The group fuzzy decision-making stage, the steps are as follows: 1. Develop detailed performance indicator evaluation content and method In the performance indicator aspect, a performance indicator is a balanced project, which represents the achievement level of a goal. But for wait. The target of the four major rings, such as the weaving and the layer, should be counted by the district quantitative value. The score of the score is composed of the scores. 2009 195524 The different evaluators have different perceptions, and the deviations of the assessments are generated, so V f ', must discuss the data collection method for each performance indicator, evaluation of the early position i, evaluation methods and standards. In this step, the knowledge accumulated during the introduction of the target structure by the bead and the dual-legged ERP contributes to the performance: tt, the establishment of the evaluation method and standards. A standard questionnaire may be developed to assist the evaluator in the assessment. in the mean time -- . To discuss and determine the weight of each performance indicator. The weight of the performance indicator can be given subjectively by the group members, or by comparing the two indicators. ❿ 2. 疋, 疋, for example, using the ahp (anaiytic hierarchy process) method (Saaty ' 1980) to calculate the relative weight of the indicator. Conducting fuzzy group performance evaluations Because some performance indicators cannot obtain a clear numerical quantified representation, the semantics (linguis1; ic) method will be used for evaluation. The described semantic evaluation value will be converted into the corresponding fuzzy number. According to the nature of the fuzzy number and the expansion principle (Iq ir and Yuan. 1995), the false s and two regular two-angle fuzzy numbers (N〇rmai triangUiar fuzzy Numbe 〇r) g = (~0',c]) and 5 = , then its fuzzy algebra operation is as follows: (1) Right ten Yin 2 = (a, + a2, h + 62, 4 + c2) (2) 7;®f2 =(α, χα2,Ζ>, xb2,c] xc2) The evaluator uses the collected data to express the meaning of the performance indicator with the set of meaning values of fork={^P, P, ^G, FG}. Weight value evaluation, where the performance semantic value is VP = Very Pool, L = Pool, F = Fair, G = Good, and Very Good. The corresponding three-dimensional fuzzy number corresponding to it is shown in the following table. The semantic weight value can also be determined by the group's common 11 200945224 -, where the performance semantic value is VL = Very Low, L = • Difference () M k (Median), Η = High (High), VH = very high (Very Η _) ' and the semantic weight values may include VL = very poor (v town l〇w), l = difference (Low), Μ = normal (Median), H = high (High), VH = Very high (Very 耵gh); the corresponding triangular fuzzy number is shown in the following table i: semantic weight value and semantic evaluation value to triangular fuzzy number table 语 semantic weight value __ ·. very low (VL) low (L) medium (M) High (H) Very high (VH) Semantic evaluation value is poor (VP) Difference (P) Normal (F) Good (G) Very good (VG) Fuzzy number (0,0,2.5) (0, 2.5 , 5.0) (2.5, 5.0, 7.5) (5.0, 7.5, 10.0) (7.5, 10.0, 10.0) In addition, if the evaluator has a different range for the fuzzy value corresponding to the meaning conversion of this language, the system also It can be set by the evaluator. It is added that the 'fuzzy set theory' uses the value of the "membership function" to describe the trait of a concept, that is, the value between 0 and 1 is used to indicate that an element belongs to a certain concept. Degree, this value is called the membership degree of the element to the set, and the commonly used fuzzy number type includes LR fuzzy number, trapezoidal fuzzy mimber and triangular fuzzy number. For ease of calculation, the present invention will summarize the semantic evaluation values of the Efficient fuzzy weighted average algorithm proposed by Lee and Park (19 97) to obtain an Aggregated fuzzy rating and the entire system. Total fuzzy assessment of performance 12 200945224 • (Total fuzzy rating). (II) Analysis of the results of the evaluation and the stage of system improvement, the steps are as follows: 1. Analysis of the results of the evaluation. First, the present invention will use the Fuzzy total integral value method proposed by Liou and Wang (1992) to The comprehensive fuzzy evaluation value of the facet and the total fuzzy evaluation value of the overall system performance are converted into the semantic value. The reason for using this fuzzy integral number method is that and Wre (2001a, 2001b) will use four important mathematical ordering methods published in the next two years, and use nine main mathematical basic theorems to verify and find fuzzy integrals. The number ordering method conforms to the eight theorems and is the best of all fuzzy sorting methods. Suppose d is a fuzzy number, and the peak value 1 of its membership function is divided into left and right parts. Then the left membership function is defined as the slice and the right membership function is defined as eight, and #—[(U] , // ~>[〇, 1] Definition 4 is the inverse of the slice and & is the inverse of //. Then the left integral val ue is defined as: right The right integral value is defined as: Integrate the left and right integral values with the optimistic parameter β to obtain the total product of 2 200945224 score (total integral value)
Ir (2) = eiR (A) + (1- 0yi (2), Θ 6 [0,1] m ❹ 其中，樂觀參數夕值可 須針對此參數值進行探 .」與「事後分析」兩種 「事前分析法」即根據 意評估值加以彙總，利 。此方法只依據決策者 進行其他的判斷或調整 策者依據決策時對於環 —個主觀的參數值。本 料之主要構面維度推導 之公式，以作為個案研 用’並和評估者以事後 行比較。由於每一個主 個系統績效的總模糊評 確值（crisp value)，本發 ce score)。因此可藉由 標構面中每一個主要構 。由於語意值為人們較 將績效分數轉換成為語 。並由各績效指標的語 觀察各主要構面和整體 分數轉換為語意值之方 能影響 討。本 方式討 所有決 用公式 所做的 。而「 境與判 發明將 適合的 究時進 分析法 要構面 估值所 明將其 此量化 面和整 為熟悉 意值之 意評估 ERP系 法說明 最終評估結 發明將分成 論此樂觀參 策者的語意 推導而成一 語意值直接 事後分析法 斷的樂觀程 依據所收集 事前分析之 行語意評估 所得之樂觀 的综合模糊 求得的全積 視為績效分 的績效分數 個系統的績 的栺述方式 4田述，以利 值結果隨著 統績致之變 士口下： 果’因此必 「事前分析 數夕。所謂 權重值和語 1固參考數值 計算’不再 」則是由決 度’協商出 個案企業資 樂觀參數夕 值彙總時使 參數Θ值進 評估值和整 分值均為明 數（performan 觀察績效指 效變化情形 ’因此亦可 管理者瞭解 時間推移， 化。其績效 200945224 * 假设5吾思值集合為= {ΚΡ5 P，p ; 定義〜=R，...，3p}， 之排序為 其中’ i 且由低而高遞增排列，p表示人 設定的語意值個數，則所有的g全積分值’、σ叉中所 Ιτ«) < I丁(d:、< …< I丁(jjρ) 〇 則 其全積分值為 ：Ir(r)， /,(<)</,(r)</r(^+l)。 假設#為 、 ’r di) 等數值 的差距之 轉換 為语意值的法 則為： (1) 若Af二 A· ’則其 值為孑。 (2) 若Af = IT(r)~Ir(dl+] )| ，則其 值為。 :+1 。 (3) 若Μ = 城Μ[/Γβ)+/, ❹Ir (2) = eiR (A) + (1 - 0yi (2), Θ 6 [0,1] m ❹ where optimistic parameter eve value can be probed for this parameter value." and "post-mortem analysis" The “ex ante analysis method” is based on the evaluation value, which is based on the decision-maker's other judgments or the adjustment of the policy-based decision-making value for the ring-subjective parameter. The main facet dimension of the material is derived. The formula is used as a case study' and compared with the evaluator after the fact. Because of the total fuzzy value of each main system performance, the ce score. Therefore, each of the main structures in the surface can be constructed. Because the semantic value is more people convert the performance score into words. And the performance of each performance indicator to observe the major facet and the overall score can be converted into semantic values. This method discusses all the formulas used. And the "evaluation of the situation and the judgment of the invention will be suitable for the face-to-face analysis. The face-to-face valuation will be based on the evaluation of the ERP system. The final evaluation will be divided into the optimistic participants. The meaning of the semantic derivation is a lexical value. The optimistic process of the post-analysis method is based on the optimistic comprehensive fuzzy ambiguity obtained from the evaluation of the pre-existing analysis. 4 Tian Shu, the result of the profit value is changed with the general score: "Therefore, it must be analyzed before the eve. The so-called weight value and the language 1 solid reference value calculation is no longer "consulted by the decision" When the case value of the optimistic parameter of the case is summarized, the parameter Θ value is added to the evaluation value and the whole value is the explicit number (performan observes the performance change effect situation), so the manager can understand the time lapse. The performance is 200945224 * Assumption 5 The value set is = {ΚΡ5 P,p ; the definition is ~=R,...,3p}, the order is as 'i and is arranged by low and high increments, p is the language set by the person For the number of values, all g full integral values ', στ in the σ fork«) < I D (d:, < ... < I D (jjρ) 其 its full integral value: Ir (r) , /, (<)</,(r)</r(^+l). Assume that the difference between the values of #, , 'r di) and the value of the semantic value is: (1) If Af Two A·' then its value is 孑. (2) If Af = IT(r)~Ir(dl+] )| , then its value is . :+1. (3) If Μ = Μ[/Γβ)+/, ❹
Q 2. 即 則其模糊综合評比值F 的語意值為介於^ 和g之間。 由此語意轉換法則可將每一主要 啤杜姑·ί私， 資構面的綜合模糊 ° ^個系統績效的總模糊評估值所求得之績效 分數轉換成為語意值。 、.責效 擬定系統改善方案 =上依據所評估之每一主要構面的综合模糊評估值和 整個系統績效的總模糊評估值得到的解模糊化 (defuzzy)的績效分數和所轉換的語意值結果，將可依 據時間的推移產生各主要構面和整體系統績效之變化 趨勢。決策者將可選定時間區段晝出各主要目標、各 主要構面和ERP全系統之績效分數曲線圖。根據這些 15 200945224 - 績效評估報表，管理者可以進而分析那些目 / 度較高？那些目標仍有待努力？然後檢討尚 標的績效指標，擬定具體的改善與發展方案 ERP系統將隨著企業的成長與企業環境的變 步地發展或更新，因此ERP系統績效評估系 與時倶進。週期性的ERP系統績效評估是必 例行工作，以提供管理者擬定改善策略的參 任何新的評估績效指標或績效指標構面必須 ，則依循本架構之步驟即可達成。 〇 至於所述績效評估報表最終所得分數及語意 其具體呈現形式可為多種，例如：樹狀表現形式 式、圖形表現形式等；所有目標的績效分數可以 紅燈儀表板方式顯不’又績效的檢視可為早期檢 亦可為多期檢視態樣者。 接著，就本發明所揭運用模糊理論評估企業 績效評估系統面而言，其構成如第3圖所示，係 〇 績效評估指標資料建構單元，係藉以提供輸 、調整績效評估指標相關資料條件； . 績效評估指標資料儲存單元，係藉以儲存由 評估指標資料建構單元所輸入、調整之績效評估 ，進以形成若干績效指標構面，所述績效指標構 為四大範疇，此部份於前文已有說明，即不重複 群體模糊運算單元，係可依據前述績效指標 進行局部或全部構面之模糊評估計算，以獲得一 值，所述模糊評估值可為分數值或語意值； 標的達成 未達成目 。由於 化而進一 統也必須 須執行的 考。若有 進行修正 值結果， 、表格形 綠、黃、 視態樣、 資源規劃 包括： 入、建制 前述績效 指標資料 面可區分 贅述； 構面條件 模糊評估 16 200945224 評估結果輸出單元，藉以將前述模糊評估值具體 呈現者；其中所述模糊評估值可透過電腦螢幕顯示或 輸出方式具體呈現，而+ 斗、丄 兄 而王現之形式如刖文所列舉。Q 2. That is, the semantic value of the fuzzy comprehensive evaluation value F is between ^ and g. The semantic conversion rule can convert the performance scores obtained by the total fuzzy evaluation value of each major beer Du Gu· 私 , 、 、 、 、 、 、 、 、 、 、 、 、 、 、 、 , blame the system improvement plan = the defuzzy performance score and the converted semantic value based on the comprehensive fuzzy evaluation value of each major facet evaluated and the total fuzzy evaluation value of the overall system performance As a result, trends in the performance of each major facet and overall system performance will be generated over time. The decision maker will be able to select a time segment for each major target, each major facet, and the ERP system-wide performance score graph. Based on these 15 200945224 - Performance Evaluation Reports, can managers further analyze those goals? Those goals still have to be worked hard? Then review the performance indicators that are still in place and formulate specific improvement and development plans. The ERP system will be developed or updated as the company grows and the business environment changes. Therefore, the ERP system performance evaluation system is timely. Periodic ERP system performance assessment is a mandatory task to provide managers with the ability to develop an improvement strategy. Any new assessment of performance indicators or performance indicators must be achieved by following the steps in this architecture. As for the final scores and semantics of the performance evaluation report, the specific presentation forms can be various, such as: tree expression form, graphic expression form, etc.; performance scores of all targets can be displayed in a red light dashboard mode. The review can be an early detection or a multi-period review. Then, as for the application of the fuzzy theory evaluation enterprise performance evaluation system disclosed in the present invention, the composition thereof is as shown in FIG. 3, and the performance evaluation index data construction unit is used to provide relevant data conditions for the transmission and adjustment performance evaluation indicators; The performance appraisal indicator data storage unit is used to store the performance appraisal input and adjustments by the evaluation index data construction unit, and to form a number of performance indicator facets. The performance indicators are organized into four categories. It is noted that the group fuzzy operation unit is not repeated, and the fuzzy evaluation calculation of the local or all facets may be performed according to the foregoing performance indicators to obtain a value, and the fuzzy evaluation value may be a fractional value or a semantic value; Head. As a result of the advancement of the system, it must also be tested. If there are correction results, tabular green, yellow, visual, and resource planning, including: Entering and establishing the above-mentioned performance indicators, the data surface can be distinguished; The surface condition fuzzy evaluation 16 200945224 The evaluation result output unit, thereby blurring the foregoing The evaluation value is specifically presented; wherein the fuzzy evaluation value can be specifically presented through a computer screen display or output mode, and the form of the + bucket, the brother and the king is as listed in the text.
本發明之優點I 業I以自行決定建立其目標與績效指標架構， 業田初‘入ERp系統之目標或後來所選定之目 设計其評估之構面 兼顧定量與定性指 體評估決策方式給 識，進而獲得更加 上述實施例所掠|千# α 跄、悉、Μ 4* — 厅揭不者係糟以具體說明本發明， 利範圍；孰来士 丁說明，當不能以此限定本發 ^ Si m Ρ1, A'之人士當可在瞭解本發 等變f盥I a , /、乜改而達到4效之目的 导灸更與修改，皆應涵芸 定範_中。 ；后所述之申請專利範 ❹【圖式簡單說明】 本發明所揭評估大、+ τ拓方法之文字方塊圖。 本發明之系統流栽〜一 β &•文子方塊簡示圖。 本發明企業資源規劏么&冰Λ、^ _向 %劃糸統構成簡示圖 【主要元件符號說明】 # :無元件符號 輸出 紙本 依據 標加 用集 觀意 文中 之專 之精 而此 所界 標’使績效指標完整。並且採 定評估值，減少評估者個人主 客觀準確而實用的評估結果。Advantages of the Invention I Industry I establish its target and performance indicator structure at its own discretion, and the objective of the ERP system or the selected content of the design of the ERp system is designed to take into account the quantitative and qualitative reference evaluation methods. Knowledge, and then obtain more of the above-mentioned embodiments Sweep|千#α 跄, Μ, Μ 4* — 厅 揭 揭 系 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以 以^ Si m Ρ1, A's people can understand the changes in the hair, such as the change of f盥I a, /, tampering to achieve the purpose of 4 effects, and should be revised. The patent application model described later [simplified description of the drawing] The text block diagram of the large and + τ extension method is disclosed in the present invention. The system of the present invention flows a ~-β &• text sub-block diagram. The enterprise resource specification of the present invention & hail, ^ _ to % 糸 constituting a simplified diagram [main component symbol description] #: no component symbol output paper according to the standard plus the use of the collection of the essence of the essence This landmark 'completes performance indicators. And the evaluation value is used to reduce the evaluator's personal objective and accurate and practical evaluation results.
第 第 第 圖 圖 圖 17First Figure Figure 17