Research And Application Of Rough Set-based Business Intelligence Decision

Business intelligence refers to the vast amounts of business data from the data mining technology to extract and discover knowledge of commercial operation of the rules used to guide the way for further business decisions is the core competitiveness of modern enterprises to obtain an important means. For a national chain of mobile phone sales companies, sales outlets of choice are frequent and important in their daily business decision-making. However, the choice of sales outlets, including regional consumer facing features, user age structure, brand image, and the interaction of many factors, making the simple decision-making difficult thoughtful man, and ultimately affect the overall performance of enterprise sales and development network.In this paper, a business intelligence decision-making information systems, sales outlets through the selection of case studies, the use of rough set theory, extraction and sales outlets found in the knowledge of decision-making rules, as a sales network of intelligent decision-making basis, to assist to enhance business intelligence decision-making level. We build business intelligence decision system consists of two subsystems:subsystem generated rules and intelligent decision-making subsystem. In the rule generation subsystem, the choice of sample cases and cleaning module is a clear case of the sample, its reference to pre-quantitative data to construct the sample structure; decision table module is structured by the decomposition of the sample case data according to the evaluation system factors, build decision-making table, and the attribute reduction and value reduction; rule extraction module is the knowledge of the rules found in the rough set algorithm, calculation of the factors that affect the decision making of the contribution of sales outlets, sales of extraction knowledge of network decision-making rules; Rules extracted from the database management are responsible for knowledge of the rules into the rule base to the table of decision rules for sales outlets to provide knowledge to support intelligent decision-making subsystem. Intelligent decision making in the subsystem, data acquisition and preprocessing module decision-making on the current sales outlets for the initial data collection, recording and pretreatment; rule-based reasoning and analysis module is based on initial data, in the knowledge of the rules in the rule base compare the collection analysis and found that the best match of the decision-making rules; Business decision-making module will form the final policy recommendations, decision support business intelligence implementation.Generation in decision-making rules, the paper based on rough set theory, through the existing extensive experience in sales outlets selected data analysis and pre-established factors of discrete decision table. Then use simple algorithms to calculate factors for each property and sales outlets condition attribute set dependence, factors that influence the properties of the relationship between attributes and decision-making to assess the various factors associated with decision-making level, discover and extract useful features set of attributes and characteristics of properties(nuclear), and the impact of small degrees by removing redundant factors or to obtain a streamlined decision-making knowledge for sales outlets rule.Finally, business intelligence decision system established by a national chain of large mobile sales application for a trial on the real data analysis and extraction, found a number of decisions rules and making the results of sales outlets, validate the results of the system.