Fuzzy Multiple Attribute Behavior Decision Making Methods And The Application In E-commerce Recommendation

We make various decisions in our daily life. We make decisions on every occasion: from casual ones such as what to eat for lunch to more serious decisions such as an individual's future course and government policy. Decision making,therefore, becomes a rather important concept when considering practical activities and making choices in life. In some decision problems, there are multiple attributes,which are difficult to be measured quantitatively, instead, the attributes are described with fuzzy linguistic variables. Meanwhile, the psychological preference from the decision maker plays an important role in decision process. Therefore, the paper is focused on fuzzy multiple attribute behavioral decision making methods, which include fuzzy linguistic information processing and transforming, behavioral decision methods under linguistic evaluations, behavioral MADM methods based on prospect theory and three-way behavioral decision methods based on prospect theory. The main contents are as follows.(1) Fuzzy linguistic information processing and transforming. Sometimes, people are prone to describe objects with natural languages including words and sentences,so it is useful to compute with words. However, it is obvious that words mean different to different people. Hence, linguistic information must be transformed into numerical forms before aggregation. To deal with this problem, we propose a novel transforming method based on sample survey. Firstly, through survey questionnaire,we collect the numerical data corresponding to each word. Then, we preprocess the collected data to remove those invalid or unreasonable points. Meanwhile, fuzzy sets are valid to present uncertainty, and the triangular fuzzy sets are the simplest one,which can capture the similarity and dissimilarity to the same word from different people. Therefore, based on means and deviations of the remaining data, we ultimately encode the linguistic terms into triangular fuzzy sets and establish the codebooks.(2) Behavioral decision methods under linguistic evaluations. In some decision making problems under risk, people prefer to assess the alternatives qualitatively instead of quantitatively. Meanwhile, the psychological attitudes of the decision makers play an important role. Thus, in this paper, we provide a prospect theory based method for linguistic decision making under risk. Primarily, we transform linguistic terms into triangular fuzzy numbers, such that linguistic evaluations can be changed into numerical forms. Then, with the aid of prospect theory, we can compute the probability weight functions and the linguistic value functions, based on which the prospect values of the alternatives are obtained. Finally, the alternatives are ranked in descending order with respect to the prospect values, and the first one is the optimal choice.(3) Behavioral MADM methods based on prospect theory. In MADM problems,the attributes are often rated in linguistic variables, some researchers transform them into numerical values through formulas. However, it might be inconsistent with real human thinking in some extent. In order to deal with such problems, a sample survey based MADM method with prospect theory is proposed. Firstly,through sampling survey, we collect the data in single point format corresponding to the words, and establish the codebook by mapping words into fuzzy sets, after that fuzzy operation rules are suitable for them. Secondly,based on the reference points, the gain and/or loss matrix is obtained, in accordance with the value function, the prospect value matrix is constructed. Finally,the alternatives are ranked in descending order respect to the defuzzified values. The first alternative is chosen as the best decision result.(4) Three-way behavioral decision methods based on prospect theory. The three-way decision method is a kind of new developed uncertain decision theory. The method divides a set of decision alternatives into three regions, called the acceptance,rejection and uncertain regions constructed from a pair of thresholds. However,seldom researches about three-way decisions take psychological attitudes and preferences of the decision makers into consideration. In this paper, we propose a novel three-way group decision method, which considers the psychological attitudes in three-way decision method. Firstly, we provide the concept of three-way group decisions. The three-way approximation is constructed by many decision makers with respect to a single critical value. The objects are still divided into acceptance,rejection and uncertain region, respectively. Then, for the proper region (uncertain region, in general), we propose a decision method, which considers the psychological preferences through fusing prospect theory into three-way group decision methods.The alternatives in the proper region are ranked in accordance with the weighted prospect values. Finally, the optimal choices are made.With the development of the E-commerce technology, the E-commerce recommendation is more and more important. So, the feasibility and effectiveness of the proposed behavioral decision methods are illustrated through the applications in E-commerce recommendation from real life.