Font Size: a A A

Evaluation Method Of Customer Relationship Management Capability Based On BP Algorithm

Posted on:2015-09-28Degree:MasterType:Thesis
Country:ChinaCandidate:Q JiFull Text:PDF
GTID:2298330467485601Subject:Measuring and Testing Technology and Instruments
Abstract/Summary:PDF Full Text Request
As the productivity keeps increasing, the shortage of the production capacity and the commodities of the whole socity is eventually overcome. The customer has replaced the role that products plays and becomes the predominance. Under this circumstance, the importance of Customer Relationship Management (CRM) is increased and the theory of CRM is improved gradually. However, many companies are not able to implement CRM so well that they need to actively evaluate their ability in customer relationship management to overcome their own shortcomings, and finally, to improve the companies’core competitiveness.Four key indicators are created from the CRM process model:customer knowledge, customer interaction, customer value, and customer satisfaction. Several second level indicators are built based on these four key indicators. One hundard companies’data is gathered according to the evaluation indicators. Coefficient of variation and Simple linear weighted function are used to calculate the indicators’weight and the score of the CRM ability, respectively. The values of action in decision table for management are thus obtained. Then the theories related to Rough Set are used to perform data preprocessing for the table. Continuous attributes are discretized by equal-frequency discretization algorithm. Redundant attributes are eliminated by Pawlak signatures based algorithm for knowledge reduction. The optimal sets of properties are thus obtained and the dimension of BP neural network modeling is decreased. The neuron number of hidden layer is calculated by the empirical formulas. An evaluation network model of CRM ability is then created and trained by BP neural network with the sample data. Twenty test samples is used to analyze and verifiy the model.A CRM system is built based on the CRM theory and the study of CRM ability evaluation. The system’s overall atchitecture, the implementation of features, the design of database and the examples of implementation are imtroduced. The evaluation model is applied to the system, perform live assessment for the company’s CRM ability, and adjust the mode of company’s own management. The study of the thesis becomes more meanful and pratical.
Keywords/Search Tags:CRM, BP Neural Network, Rough Set, CRM Ability Evaluation, Pawlaksignature
PDF Full Text Request
Related items