The Design And Implementation Of Business Intelligence Analysis System

Communication in today’s society has become a major activity, from wiredcommunication previous years to wireless mobile communication present. The difficulty ofpeople to participate in commercial activities has also decreased. In this context, the data ofbusiness areas is also rapidly expanding, the data can be able to analyze manually in the pastis now becoming more and more difficult. So now business intelligence as role to supportbusiness is becoming increasingly apparent.With web standards and industry specifications have been proposed, more and morebrowser vendors is complying with industry standards, so standard software once written canwork on all current operating systems and all devices. For example, desktop, Android andApple phones, pad, laptops can be partial or complete support without additional installationof software to run the application. Existing browser standards can perfectly present variousdata graph which could only rely on plug-in in the past. And given the standard browser richimage features that help splitting data with demonstrate, the server is only responsible forextracting data, the client is responsible for the data graphically, we know that the graphics areprocessor-cost computing time. It also makes the software closer to a traditional operatingsystem-based software.Business intelligence has many hot research field. Community detection, productrecommendations etc. are important in today’s computer applications research. Manyalgorithms in these areas are from data mining, statistical, business and other disciplines.These algorithms is usually able to find information that cannot be found in the analysis, theeffect of the algorithm itself define a good heuristic function, the direction of process inalgorithm is to heuristic function optimization. These algorithms get results throughcontinuous iteration, such as k-means algorithm; Or using a special data structure to optimizetime efficiency, such as FP growth tree algorithm; Or probability to fit through the uncertaintyof the real world, such as the ant colony algorithm.In this paper, by drawing on the classical theory of data mining and business intelligencerelated literature, in the foundation that have a more comprehensive understanding of businessintelligence, has developed an algorithm based on classic business intelligence analysissystem the system will be deconstructed into five business processes category: customeranalysis, market analysis, revenue analysis, operational monitoring, policy support. Usingadvanced imaging technology, fully learned the needs of business field about information andclassic algorithm used in business field data about decision tree algorithm, improved antcolony algorithm, modified k-means algorithm. Analysis of business data from network and history indicate that the accumulation ofbusiness intelligence analysis system can effectively support most of the information youneed. It can help companies to distinguish their customer s base and levels, help assessunfamiliar customer credit; can predict future market expectations of products; similarproducts can monitor sales and other business statistics digital; support optimal businesslocation selected industry; help users product recommendations.Business intelligence in the future will operate multi-level data, large-scale data, richdata structures. Therefore, the demand for a new algorithm for the new era of business,require the system to handle concurrent data online, offline processing large data cannotreduce accuracy. This is the future work direction of parallel processing needs of thealgorithm the bus data transfer and other new features.