Crm Systems, Data Mining Algorithms And Applied Research

With the wide application of CRM system, enterprise could unfold activities to meet the demand of their customers and provide them with the right products and services. After the running of CRM systems for some time, however, the enterprise may have enormous amount of customer related data. This poses the question for decision makers: How to convert large volume of data into valuable knowledge? How to use the knowledge to guide enterprise activities? Those are big challenges.So, this paper does 2 different but closely coupled research to answer those questions. The first one is about how to effectively find useful knowledge among the data, knowledge about customer classification to be exactly. The second research is about how to draw on the knowledge found to guide enterprise in their customer relationship improvement plan and the plan can maximize the enterprise profit.Customer classification is one of the most important research areas of Data Mining. It extracts data type model from huge amount of data, and the model can be reused in later customer classification and decision-making. In order to draw on the merits of different algorithms, this paper put forwards an algorithm called GA-DFC (Genetic Algorithm based Decision Forest parallel Combination algorithm) based on the thought of combination algorithm. The algorithm model contains 2 parts: on the upper part is the decision forest consists of multi-decision trees. On the lower part is the weighted network to combine the classification result of decision forest. By analysis and test, this method is proved to be more accurate than traditional method and can handle data with noise and large attribute set. The algorithm can be put into CRM analysis.Most data mining algorithms and tools stop at discovered customer models, producing distribution information on customer profiles. Such techniques, when applied to CRM problems, are useful in pointing out customers' attributes but they require human experts to process the mined information manually and they do not directly suggest actions of improving customer relationship, and lead to an increase in profit. This paper present a novel algorithm called DT-ACTsM (Decision-tree based Activities Miningalgorithm) that suggest actions to change customers from an undesired status to a desired one while maximizing objective function: the expected net profit. We also ponder these algorithms under resource constraints.This paper contains 2 consecutive and closely coupled works, which can realize the mining process from data to knowledge, from knowledge to activities, and ultimately enterprise profits.Major innovations in this paper:(1) Propose an algorithm for decision forest growing.(2) Put forward a GA based decision forest combination algorithm (GA-DFC), which is proved accurate in classification problems.(3) Give out an algorithm named DT-ACTsM that can be useful in the customer relationship improvement activities.In the era of e-business, data mining will play more and more important role and will bring CRM even forwards furthermore. So the research work of this paper has enormous theoretical and practical significance.