Fuzzy knowledge management advisor system based on computing with words technique

Today, appropriate management of intellectual capital is one of the most important concerns of successful organizations. Organizations need to identify and manage their intellectual capital in order to gain sustainable competitive advantage. Knowledge Management (KM) means doing whatever is necessary to get more utilization of knowledge resources. One of the important parts of implementing KM is proper recognition of suitable solutions according to organization’s properties and conditions. Our approach is based on Fernandez and Sabherwal contingency theory for identifying KM processes and solutions. It is difficult to quantitatively evaluate contingency factors affecting KM processes as such factors involve human perceptual interpretation with certain subjectively, uncertainty and imprecision. In this paper, we introduce a fuzzy method to find out the best solution for KM development based on computing with words technique. After collecting information from ۱۰۱ employees of different companies, we achieved the FOU (Footprint Of Uncertainty) for ۳۲ prefix or suffix of words that people used to describe factors and used them to establish the codebook; since human beings understand and express themselves naturally using ‘words’. The encoder transforms linguistic perceptions into interval type-۲ fuzzy sets (IT۲ FSs) that activate engine. We used linguistic weighted average in fuzzy inference engine for aggregation of the evaluation of different aspects and then, Karnik & Mendel and Jaccard similarity algorithm is used to produce the results. In this method, the distinction between different ideas are made with more accuracy by using linguistic variables. It overcomes the problems of modeling uncertainty during computing with words by using FOUs. The decoder maps the output of engine back into a word. The output of our model could be descriptive, comparative or multilevel in case of the maturity model. An empirical study is performed to demonstrate the implementation process. This study shows people have better understanding of method when they use their own words and the result is more accurate model of their thinking. The result of this paper is a fuzzy perceptual computer that overcomes ambiguities and all problems deal with use of a limited number of options based on Likert scale to linguistic term.