The Research On Network Market Of Educational Software Based On Data Mining

With the time passing by and the development of modern science and technology, educationalsoftware industry continuously updates its marketing model. The electronical market platform ofeducational software were set up very late and it went through a difficult initial stage from2002to2006. In modern E-commerce stage, the B2B and B2C E-commerce net tend to be the extremecompetition platform for many companies. Many companies face various circumstances, such asMarketing platform transformation, the worker weakness of advanced technology, the connectingshortage between the industry and the network environment. Because of the imperfect analysisand prediction on network market, many consumers information collecting and process methodsare not correspond with the objective realities, and that is the key to lead the bad sales. In the caseof lack of consumer in this market, and its lack of good operation strategy, and so on, all of thesedisadvantage may lead to bad circle for the development of network educational software industry.So, based on Data Mining Technology, we may do some tentative work on network educationalsoftware market analysis.This research tries to explore the new strategy in information analysis area. There are three partsin this research: The first part try to find out the highest value consumers of network educationalsoftware product. Whether in the visual store or the network invisual store, every company mustcontrol the important consumer information, then the system for collecting important consumerinformation needs efficient collection strategy, and goes through the efficient strategy to obtain thevaluable information. According to the characteristic of network educational software, industrycan also make the good collection strategy. The first step of research gives the analysis ideas anduses the RFM consumer predict model to respectively define the weighing proportion of networkeducational software product, to use this three index to find out the highest value consumer, thento explore the potential regular based on this step. The second step is based on the analysis ofbrowse and purchase path of high value consumer. We reconstruct a new path strategy ofconsumer, take correspond analysis and calculate on it. At first, we point out the inadequacies oforder traversal path model, research on the reconstruction of the previous model, then give theimproved method that adds trace-back path to order traversal path. By the TFTA algorithm, wefind out the key back tracing point in the path and grasp the time interval to analyze quantitatively.Based on this, the websites interested in by the networks customers are known exactly, which willhelp the network companies further analyze the customers’ information and draw up the correctstrategies of website design and Internet marketing. The third step is to put the algorithm intoelectronic process. We may design the process and compile the program, after that, put theconsumer data into the process to calculate and output the result. This process can improve thefeasibility and accuracy level, and it can achieve the high efficiency when the large amount ofconsumer dates are take into account. This research supports the previous academic conclusionand obeys the regular of concrete method to solve the concrete question. In the end, the authorgives the result for this research and suggests future direction, the author wishes that the thesiswould help and deepen the correlation research.