A possibilistic-valued multi-criteria decision-making support for marketing activities in e-commerce: Feedback Based Diagnosis System

Abstract

In this paper we propose a multi-criteria decision making support system, called a ''Feedback Based Diagnosis System" (FBDS), to aid the marketing team of an e-commerce (EC) organisation in its activities. The FBDS database is composed of customers' satisfaction measures. These measures are related to the different services an EC offers to its customers. Thus, they constitute a multi-criteria (MC) evaluation of EC performances. In the general framework of recommender systems, these available MC evaluations are considered as useful information for other customers to help them to objectively, rationally and exhaustively assess and compare the numerous ECs among the ones likely to meet their needs. Our FBDS is not concerned with improving or automating such a recommendation process for customers. Indeed, it is merely EC management team oriented. In fact, the MC feedback database is used to diagnose the EC health and improve its strategy. In the proposed FBDS, a possibilistic framework is combined with the multi criteria representation to capture the variability and the divergence of customers' evaluations w.r.t. each criterion. Then, an aggregation based on a weighted arithmetic mean (WAM) is proposed to obtain a synthetic appraisal of ECs. The WAM aggregation models the strategy agreed on by the EC management team. Computing the synthesis score of an EC consists in propagating the uncertainty related to its partial scores through the WAM. The possibilistic representation guarantees that no information is lost in the collective evaluation process by the consumers' community. However, diagnosis indicators are finally proposed to the marketing team to make the interpretation of some possibilistic results more comprehensive when necessary.

Data Reliability Assessment in a Data Warehouse Opened on the Web

“…Our approach is different, as we do not rely on users but rather on information about the Web data table origins to compute a reliability estimations. Among methods proposing solutions to evaluate trust or data quality in web applications, the method presented in [3] is close to the method presented in the paper. It uses possibility theory evidence theory, whereas we base our method on evidence theory.…”

“…It uses possibility theory evidence theory, whereas we base our method on evidence theory. Another difference is that in our approach global information is obtained by a fusion of multiple uncertainty models, while in [3] global information results from the propagation of uncertainty models through a aggregation function. Each method has its pro and cons: it is easier to integrate interactions between criteria in aggregation functions, while it is easier to retrieve explanations of the final result in our approach.…”

Evaluating Data Reliability: An Evidential Answer with Application to a Web-Enabled Data Warehouse

“…Among methods proposing solutions to evaluate trust or data quality in web applications, the method presented in [13] for recommendation systems is close to our proposal, but uses possibility theory as a basis for evaluations rather than belief functions. Another difference between this approach and ours is that global information is not obtained by a fusion of multiple uncertainty models, but by the propagation of uncertain criteria through an aggregation function (e.g., a weighted mean).…”

Criteria Weighting and 4P's Planning in Marketing Using a Fuzzy Metric Distance and AHP Hybrid Method

“…all the sub-criteria that can be effective in this decision process will not be included in this stage of the study and all the criteria are supposed to be independent. In the literature, during the last years, many studies have been published about marketing strategy selection and strategic marketing; by Rekik et al, proposing a multi-criteria decision making support system to aid the marketing strategy selection in e-commerce 6 , by Tsai et al, proposing an integrated model for evaluating airlines' websites effectiveness which is based on the perspectives of ''marketing mix 4Ps'' and ''website quality'' for the web-based marketing using the ANP 11 , by Liao, proposing a method that will guide the product development team to select the best marketing strategy by taking into account the price level and product/market segmentation 12 , and by Wang, providing a reference for planning brand marketing with a hybrid MCDM model combining the Decision Making Trial and Evaluation Laboratory (DEMATEL) with ANP and VIKOR methods 13 .…”