Research And Implementation Of CRM System Based On Data Mining For Coal Industry

The Pit-Coal-Vocation, as the foundational industry of our country, its effective supply gives guarantee for the need of state economy and social development, and also contributes greatly to our country's macroeconomy rising. Yet for the moment, the Pit-Coal-Vocation marketing be faced with some problems, such as marketing management concept lag, marketing tools single, information exchange not smooth, thereof urge the Pit-Coal-Corporation undertake from buy "product" turning to buy "service". Customer relationship management, especially related software's application, makes the corporation quickly gather,tidy and analysis customs' feedback information, in order to grip and supervise some emphases arrearage user's payment for goods reclaim condition, thus avoiding malignancy overdrawn and in arrears coal payment, at the same time through using this information in each tache, from pit coal product processing to after-service, could increase product ability of adaptation market,satisfy custom's need pertinently,according to market environment change, in time adopt corresponding step, modulate marketing policy, depress operating risk.This essay analyzes the character of data digging technology and CRM system, then based on them, combining character of data digging and special need of CRM system, emphasizes on the appliance of data digging technology in CRM system ,and with the Fujian province Coal-electricity put forward a new system structure-Virtual B/S Structure, which unites merits of realize data (sales plan,sales contract,custom payment and just-in-time stock) incorporate and operate automatically, all these predigest work, assure the veracity and coherence of data, achieve sharing data and real time; at last, in analytical CRM, synthetically handle coarseness multitude and fuzzy clustering technology, analysis and research pit coal client subdivision. First all, make decision information table concise by adopting Rough-degree of connection, then clustering, it effectively reduce the calculate capacity of cluster arithmetic; in addition, adopting fuzzy clustering method proceed customer segmentation, via threshold's adjustment process, could set eyes on data specimen and rest sort between's vincula, might even more objectively be symptomatic of real world.