Developing TOPSIS method using statistical normalization for selecting knowledge management strategies

Zadeh Sarraf, Amin; Mohaghar, Ali; Bazargani, Hossein Article Developing TOPSIS method using statistical normalization for selecting knowledge management strategies Journal of Industrial Engineering and Management (JIEM) Provided in Cooperation with: The School of Industrial, Aerospace and Audiovisual Engineering of Terrassa (ESEIAAT), Universitat Politècnica de Catalunya (UPC) Suggested Citation: Zadeh Sarraf, Amin; Mohaghar, Ali; Bazargani, Hossein (2013) : Developing TOPSIS method using statistical normalization for selecting knowledge management strategies, Journal of Industrial Engineering and Management (JIEM), ISSN 2013-0953, OmniaScience, Barcelona, Vol. 6, Iss. 4, pp. 860-875, https://doi.org/10.3926/jiem.573 This Version is available at: https://hdl.handle.net/10419/188566 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc/3.0/ Journal of Industrial Engineering and Management JIEM, 2013 – 6(4): 860-875 – Online ISSN: 2013-0953 – Print ISSN: 2013-8423 http://dx.doi.org/10.3926/jiem.573 Developing TOPSIS method using statistical normalization for selecting Knowledge management strategies Amin Zadeh Sarraf, Ali Mohaghar, Hossein Bazargani University of Tehran (Iran) asar[email protected] , [email protected] , [email protected] Received: October 2012 Accepted: July 2013 Abstract: Purpose: Numerous companies are expecting their knowledge management (KM) to be performed effectively in order to leverage and transform the knowledge into competitive advantages. However, here raises a critical issue of how companies can better evaluate and select a favorable KM strategy prior to a successful KM implementation. Design/methodology/approach: An extension of TOPSIS, a multi-attribute decision making (MADM) technique, to a group decision environment is investigated. TOPSIS is a practical and useful technique for ranking and selection of a number of externally determined alternatives through distance measures. The entropy method is often used for assessing weights in the TOPSIS method. Entropy in information theory is a criterion uses for measuring the amount of disorder represented by a discrete probability distribution. According to decrease resistance degree of employees opposite of implementing a new strategy, it seems necessary to spot all managers’ opinion. The normal distribution considered the most prominent probability distribution in statistics is used to normalize gathered data. Findings: The results of this study show that by considering 6 criteria for alternatives Evaluation, the most appropriate KM strategy to implement in our company was ‘‘Personalization’’. Research limitations/implications: In this research, there are some assumptions that might affect the accuracy of the approach such as normal distribution of sample and community. These assumptions can be changed in future work -860- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 Originality/value: This paper proposes an effective solution based on combined entropy and TOPSIS approach to help companies that need to evaluate and select KM strategies. In represented solution, opinions of all managers is gathered and normalized by using standard normal distribution and central limit theorem. Keywords: knowledge management; strategy; TOPSIS; Normal distribution; entropy 1. Introduction In the knowledge economy, it's necessary to find a way to create, share, and utilize knowledge if we want to have a competitive advantage (Desouza, 2003). Nowadays business environment has been more competitive, in such situation, many companies emphasize the importance of knowledge management (KM), and base the KM strategy on their unique resources and capabilities. According to (Kamara, Anumba & Carrillo, 2002), KM is the organizational optimization of knowledge to achieve enhanced performance through the use of various methods and techniques. Also, KM is a systemic way to manage knowledge in the organizationally specified process of acquiring, organizing, and communicating knowledge. More importantly, the effective KM largely begins with a proper KM strategy. Hence, in order to implement the KM successfully, there is a critical issue of how companies can better evaluate and select a favorable KM strategy. However, the KM strategy selection usually involves subjective and qualitative judgment. In particular, choosing KM strategies is a strategic issue (Bierly & Chakrabarti, 1996), which is restricted by resource needs, realistic support, time requirements, and conformity with expected outcomes or business purposes. In this sense, the treatment of KM strategy selection is required to handle several complex factors in a better sensible and logical manner. Thus, the KM strategy selection is a kind of multiple criteria decision-making (MCDM) problem, and requires MCDM methods to solve it appropriately. Many traditional MCDM methods are based on the additive concept along with the independence assumption, but each individual criterion is not always completely independent (Leung, Hui & Zheng, 2003).For solving the interactions among elements, entropy as a relatively new MCDM method was proposed by Shannon (Shannon, 1948). Although entropy has been used in this article, TOPSIS presented by Hwang and Yoon (Hwang & Yoon, 1981) has been utilized for evaluation of alternatives. According to decrease resistance of employees against implementing a new KM strategy in our organization, we extended TOPSIS and formed62empty decision matrixes and distribute them between all managers of organization to cooperate in decision making process by fulfilling those matrixes. After gathering these completed matrixes, we should make a single matrix as final decision matrix for applying TOPSIS and entropy. To reach this purpose we combined results of these 62 matrix. This was by using arithmetic average of every cell's value in all decision matrixes completed by -861- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 managers. In order to normalize decision matrix, normal distribution and central limit theorem has been utilized. 2. Literature survey In this section, some essentials of the entropy and the TOPSIS are briefly described. A table that contained all normalization methods is represented too. At the end of this part a brief description about Normal distribution and Central limit theorem is represented. 2.1. Entropy The entropy method can be used not only to quantitatively estimate data quantity, but also to calculate objectively the relative weight of information (Shannon, 1948). Entropy was originally intended to simplify a physical phenomenon of numerator turbulence degree or the probability scale under a specified condition. If entropy values are lower, the numerator degrees are more proportional, implying as close to perfect entropy as possible. Conversely, if entropy values are higher, the numerator degrees have a more irregular inflection. Therefore, entropy weight method was introduced to obtain the relative weight of each attribute. Additionally, in information theory, entropy can be used to measure expected information content of a certain message. Entropy in information theory is a criterion for the amount of uncertainty represented by a discrete probability distribution (Jaynes, 1957). Each attribute is assigned measured a value by each alternatives to calculate the entropy values. The entropy values for each criterion are then compared, and the relative significance levels of each other are calculated (i.e., the relative weight). Next, the entropy weight is obtained based on the appraisal matrix information, which belongs to the objective weight values. Calculation procedure for the entropy weight method has been described in part 3, evaluation framework. 2.2. TOPSIS The TOPSIS method was first developed by Hwang and Yoon (Hwang & Yoon, 1981) and ranks the alternatives according to their distances from the ideal and the negative ideal solution, i.e. the best alternative has simultaneously the shortest distance from the ideal solution and the farthest distance from the negative ideal solution. The ideal solution is identified with a hypothetical alternative that has the best values for all considered criteria whereas the negative ideal solution is identified with a hypothetical alternative that has the worst criteria values. In practice, TOPSIS has been successfully applied to solve selection/evaluation problems with a finite number of alternatives (Jee & Kang, 2000; Yong, 2006) because it is intuitive and easy to understand and implement. Furthermore, TOPSIS has a sound logic that represents the rationale of human choice (Shih, Syur & Lee, 2007) and has been proved to be one of the best methods in addressing the issue of rank reversal (Zanakis, Solomon, Wishart & Dublish, 1998).In this paper we extended TOPSIS for KM strategies selection problem because of following reasons and advantages as Shih and his cooperators did for consultant selection -862- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 problem (Shih et al., 2007). •A sound logic that represents the rational of human choice. •A scalar value that accounts for both the best and worst alternative simultaneously. •A simple computation process that can be easily programmed into a spreadsheet. •The performance measures of all alternatives on attributes can be visualized on a polyhedron, at least for any two dimensions. 2.3. Common methods of normalization for TOPSIS For MADM, a decision matrix is usually required prior to the beginning of the process. The decision matrix contains competitive alternatives row-wise, with their attributes’ ratings. Normalization is an operation to make these scores conform to or reduced to a norm or standard. To compare the alternatives on each attribute, the normalized process is usually made column-wise, and the normalized value will be a positive value between 0 and 1. In this way, computational problems, resulting from different measurements in the decision matrix, are eliminated (Yoon & Hwang, 1995). Attributes have been partitioned into three groups: benefit attributes, cost attributes, and non-monotonic attributes (Hwang & Yoon, 1981). A few common normalization methods are organized in Table 1 (Milani, Shanian, Madoliat & Nemes, 2005; Hwang & Yoon, 1981; Yoon & Hwang, 1995). These are classified as vector normalization, linear normalization and fuzzy normalization to fit real-world situations under different circumstances. Additionally, three forms for linear normalization are listed in Table 1. 2.4. Normal distribution In many applications in which some random variable X is normally distributed with mean µ and variance σ2, we will standardize X to obtain z-scores (z=(x-µ)/σ2). The distribution of the zscores is the standard normal distribution, that is, the normal distribution with a mean of zero and a variance of one. Therefore, if X complies N(µ, σ2), then Z abides by N(0,1) also (Belsom, 1992). The probability density function of the standard normal distribution is as follows: The cumulative distribution function (CDF) of a probability distribution contains the probabilities that a random variable X is less than or equal to X. The cumulative distribution function of the normal distribution is expressed as follows: -863- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 The normal distribution is considered the most prominent probability distribution in statistics. There are several reasons for this. One of them is that normal distribution is very tractable analytically, that is, a large number of results involving this distribution can be derived in explicit form (Casella & Berger, 1990). Table 1. Common methods of normalization for TOPSIS 2.5. Central limit theorem The importance of normal distribution as a model of quantitative phenomena in the natural and behavioral sciences is due to the central limit theorem. Under certain conditions (such as being independent and identically distributed with finite variance) the sum of a large number of random variables is approximately normally distributed, this is the central limit theorem. Many psychological measurements and physical phenomena (like noise) can be approximated well by the normal distribution. While the mechanisms underlying these phenomena are often unknown, the use of the normal model can be theoretically justified by assuming that many small, independent effects additively contribute to each observation. Zhonggen devoted to the study of central limit theorems and the domain of normal attraction for some random processes with sample paths in exponential spaces under metric entropy conditions (Zhonggen, 1997). Yokoyama studied on this line the functional central limit theorem and law of the iterated logarithm for stationary processes, not necessarily possessing the boundary decomposition, with applications to stationary linear processes (Yokoyama, 1995). Dedecker and Prieur proved a central limit theorem for the d-dimensional distribution function of a class of stationary sequences(Dedecker & Prieur, 2007). -864- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 3. Evaluation framework Selection of KM strategy is a kind of MCDM problem that needs multiple evaluation criteria. (Opricovic & Tzeng, 2004) represented that, MCDM problem solving is required to define evaluation criteria, find alternatives and evaluating of them in terms of criteria, apply an appropriate multi-criteria analysis method, and choose the best alternatives. Decision making is the process of defining the decision goals, gathering relevant criteria and possible alternatives, evaluating the alternatives for advantages and disadvantages, and selecting the optimal alternative (Hess & Siciliano, 1996). In this paper, a new method of normalization has been represented. This method called statistical method for normalization is so applicable when our data has been selected from a normal statistical society. In fact when judgment about some alternatives is implemented by some persons and we selected average value of each element and create decision matrix based on them and want to evaluate alternatives according to this matrix, we can use this method for normalization of decision matrix. We use in this article consists of four steps. In the next section a brief description about each of these steps has been represented. •Defining the problem objectives •Defining alternatives and criteria for evaluating •Applying ENTROPY model and TOPSIS •Choosing the most appropriate strategy 3.1. Defining the problems objectives: As mentioned in the past section, decision making is the process of defining the decision goals, gathering relevant criteria and possible alternatives, evaluating the alternatives for advantages and disadvantages, and selecting the optimal alternative (Hess & Siciliano, 1996). Each organization has it's own purpose by implementing KM strategies. For instance, KM is the way to improve an organization’s performance, productivity, and competitiveness and to promote learning, sharing, and usage of knowledge. The purpose of KM can be different such as: to initiate action based on knowledge; to support business strategy implementation; to become an intelligent enterprise; to increase competitive advantage; to create an innovative culture and environment; to entrench collaboration as a work practice; and to improve work efficiency (Plessis, 2005). In this phase, the objectives of our decision should become evident. Here is defining and choosing the appropriate KM strategy as defining the problem objectives in phase 1. -865- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 3.2. Defining alternatives and criteria for evaluating Defining alternatives and criteria for evaluating of them is the second phase as gathering relevant criteria and possible alternatives. In selecting appropriate strategy for an organization, it is critical to form a team and involve several experts from different departments to create the best adaptability between organization and proposed strategies for organization (Saremi, Mousavi & Sanayei, 2009). It's very important to make an effective and efficient communication between different experts because the better the parties are informed about strategy selection, the higher the probability that the parties will be committed to supporting this selection, The more different perspectives are initially taken into account, and the greater the complexity of convergence, the smaller the chances of addressing the wrong problem and reaching an inadequate solution (Karacapilidis, Adamides & Evangelou, 2006). The objectives of selection, the scope of selection, and the possible alternatives should become defined as well as possible. In this phase, after gathering data by Interviewing with elites, the data has been categorized, analyzed, and summarized to decision making matrix. As to alternatives of KM strategy, (Hansen, Nohria & Tierney, 1999) represented two types of KM strategies: the codification strategy (seeking to document and store knowledge in databases) and the personalization strategy (seeking to develop networks of people for communicating ideas). In our research, the strategic management team of the organization defined third strategy as Blend strategy (a mixture between codification and personalization). Evaluation criteria of KM strategy can range from top management support, communication, creativity, culture and people, sharing knowledge, incentives, time, and evaluation (Martensson, 2000). Strategic management team of the organization represented that evaluation criteria of KM strategy can range from top management support, time, cost, degree of acceptance by employees, technical knowledge, and knowledge sharing. So, 6 criteria selected for evaluation of 3 chosen strategies. At the end of this phase a set of possible alternatives for implemention in organization is prepared that we called them A = {A1, A2, …, Am}. Also, a set of necessary criteria selected that we called them C = {C1, C2, ..., Am}. 3.3. Applying entropy and TOPSIS methods After defining alternatives and evaluation criteria, it is necessary to apply an entropy model and the TOPSIS; the entropy model is used to calculate the elements of evaluation criteria's weights, and the TOPSIS is used to solve problem and choose the best strategy for this organization. Here are the steps of applying entropy and TOPSIS on this problem. -866- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 3.3.1. Establish a decision matrix for alternative performance According to select the best knowledge strategy for our organization, we formed strategic team from top managers of organization at first. Outcome of this team after forming several meeting, as mentioned in past paragraphs, was 3 strategy and 6 criteria for evaluation of them. After definition of alternatives and criteria, we must gather relevant data for these alternatives. According to this purpose, 62 empty decision matrix is included 3 rows and 6 columns formed and delivered to all managers of organization in strategic, middle and functional level. There were 62 managers in these levels in our organization. These managers evaluated different alternatives and completed these matrixes. In order to evaluation of results, we, as investigative teams, need just one matrix. As mentioned in the past paragraph that we had 62 matrixes, it was necessary to convert these matrixes to only one matrix. To reach this purpose, we formed an empty matrix by 3 rows and 6 columns and filled it's cells by using arithmetic average of that cell in all 62 matrixes by below form: Rij is the final value for each final decision matrix cells and xij is that cell value in 62 matrixes fulfilled by 62 managers. The strategy selection problem can be expressed in the matrix format for k-th decision maker as follows: Where fij is a linguistic variable, indicates the performance rating of each ith alternative with respect to each jth criterion. In fact each element of final matrix is the average of that element in 62 primary decision matrixes. We have shown our alternatives as Si = {S1, S2, S3} and criteria as Cj = {C1, C2, C3} and the data belongs them in Table 2 as final decision matrix. C1C2C3C4C5C6 S15.41 24 15,700 7.25 6.58 7.69 S28.32 12 6,400 3.5 4.12 3.87 S35.98 19 12,800 5.2 8.5 5.63 Table 2. Final decision matrix -867- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 Kamara, J.M., Anumba, C.J., & Carrillo, P.M. (2002). A clever approach to selecting a knowledge management strategy. International Journal of Project Management, 20(3), 205-211. http://www.sciencedirect.com/science/article/pii/S0263786301000709 http://dx.doi.org/10.1016/S0263-7863(01)00070-9 Karacapilidis, N., Adamides, E., & Evangelou, C. (2006). A computerized knowledge management system for the manufacturing strategy process. Computers in Industry, 57, 178-188. http://www.sciencedirect.com/science/article/pii/S0166361505001399 http://dx.doi.org/10.1016/j.compind.2005.07.001 Leung, L.C., Hui, Y.V., & Zheng, M. (2003). Analysis of compatibility between interdependent matrices in ANP. Journal of the Operational Research Society, 54(7), 758-768. http://www.palgrave-journals.com/jors/journal/v54/n7/full/2601569a.html http://dx.doi.org/10.1057/palgrave.jors.2601569 Martensson, M. (2000). A critical review of knowledge management as a management tool. Journal of Knowledge Management, 4(3), 204-216. http://www.emeraldinsight.com/journals.htm? issn=1367-3270&volume=4&issue=3&articleid=883705&show=abstract http://dx.doi.org/10.1108/13673270010350002 Milani, A.S., Shanian, A., Madoliat, R., & Nemes, J.A. (2005). The effect of normalization norms in multiple attribute decision making models: A case study in gear material selection. Structural Multidisciplinary Optimization, 29(4), 312-318. http://link.springer.com/article/10.1007/s00158-004-0473-1 http://dx.doi.org/10.1007/s00158-004-0473-1 Opricovic, S., & Tzeng, G.H. (2004). Compromise solution by MCDM methods: A comparative analysis of VIKOR and TOPSIS. European Journal of Operational Research, 156(2), 445-455. http://www.sciencedirect.com/science/article/pii/S0377221703000201 http://dx.doi.org/10.1016/S0377-2217(03)00020-1 Plessis, M. (2005). Drivers of knowledge management in the corporat environment. International Journal of Information Management, 25(3), 193-202. http://cat.inist.fr/? aModele=afficheN&cpsidt=16859501 http://dx.doi.org/10.1016/j.ijinfomgt.2004.12.001 Pratyyush, S., & Jian-Bo, Y. (1998). Multiple Criteria Decision Support in Engineering Design. Berlin: Springer Verlag. Saremi, M., Mousavi, S.F., & Sanayei, A. (2009). TQM consultant selection in SMEs with TOPSIS under fuzzy environment. Expert Systems with Applications, 36, 2742-2749. http://www.sciencedirect.com/science/article/pii/S0957417408000766 http://dx.doi.org/10.1016/j.eswa.2008.01.034 -874- Journal of Industrial Engineering and Management – http://dx.doi.org/10.3926/jiem.573 Shannon, C.E. (1948). The mathematical theory of communication. Bell System Technical Journal, 27, 379-423. http://cm.bell-labs.com/cm/ms/what/shannonday/shannon1948.pdf http://dx.doi.org/10.1002/j.1538-7305.1948.tb01338.x Shih, H.S, Syur, H.J, & Lee, E.S. (2007). An extension of TOPSIS for group decision making. Mathematical and Computer Modeling, 45, 801-813. http://www.sciencedirect.com/science/article/pii/S0895717706003025 http://dx.doi.org/10.1016/j.mcm.2006.03.023 Starr, M.K., & Zeleny, M. (1977). Multiple Criteria Decision Making. New York: North Holland Publication Company. Yokoyama, R. (1995). On the central limit theorem and law of the iterated logarithm for stationary processes with applications to linear processes. Stochastic Processes and their Applications, 59, 343-351. http://www.sciencedirect.com/science/article/pii/0304414995000389 http://dx.doi.org/10.1016/0304-4149(95)00038-9 Yong, D. (2006). Plant location selection based on fuzzy TOPSIS. International Journal of Advanced Manufacturing Technology, 28(7-8), 839-844. http://link.springer.com/article/10.1007/s00170-004-2436-5 http://dx.doi.org/10.1007/s00170-004-2436-5 Yoon, K.P., & Hwang, C.L. (1995). Multiple Attribute Decision Making: An Introduction. Thousand Oaks, CA: Sage Pub. Yoon, K., & Hwang, C.L. (1980). Multiple Attribute Decision Making Methods and Applications. A State of the Art Survey. Berlin: Springer Verlag. Zanakis, S.H., Solomon, A., Wishart, N., & Dublish, S. (1998). Multi-attribute decision making: A simulation comparison of select methods. European Journal of Operational Research, 107(3), 507-529. http://www.sciencedirect.com/science/article/pii/S0377221797001471 http://dx.doi.org/10.1016/S0377-2217(97)00147-1 Zhonggen, S. (1997). Central limit theorems for random processes with sample paths in exponential Orlicz spaces. Stochastic Processes and their Applications, 66 , 1-20. http://www.sciencedirect.com/science/article/pii/S0304414996001172 http://dx.doi.org/10.1016/S0304- 4149(96)00117-2 Journal of Industrial Engineering and Management, 2013 (www.jiem.org) Article's contents are provided on a Attribution-Non Commercial 3.0 Creative commons license. Readers are allowed to copy, distribute and communicate article's contents, provided the author's and Journal of Industrial Engineering and Management's names are included. It must not be used for commercial purposes. To see the complete license contents, please visit http://creativecommons.org/licenses/by-nc/3.0/. -875-