The Research On Very Large Scale Customer Relationship Management Strategies

At present, the world is in an era of explosive data growth. The size of data isgrowing exponentially. Massive data is difficult to manage and maintain but alsoprovides businesses with great potential value. Data is like a magic diamond mine; theprimary value was discovered and still continues to give. Its true value is like aniceberg floating in the ocean, at first you only see the tip of the iceberg, and most ofthem are hidden beneath the surface. The era of ultra-large-scale customers,companies have massive enterprise customer data on one hand and a morecomprehensive understanding of the customer needs, to provide a more abundant rawmaterial; on the other hand, with the proliferation of data-scales, traditional dataanalysis methods for massive data inadequate treatment on some customerrelationship management thinking is no longer able to effectively guide theenterprises targeted to develop scientific, timely and customer relationshipmanagement strategies.This paper analyzes the characteristics of ultra-large-scale customerenvironments, based on the traditional customer relationship management theory tosort out and explore the impact of ultra-large-scale customer environments. This is tobring enterprise customer relationship management, through comparative analysisfound that the ultra-large-scale enterprise customer relationship managementcustomer environments trends. Enterprises also need to improve customermanagement capabilities to support large scale customer data; ultra-large-scaleenterprises will help customers integrate all relevant data, customer information tobuild a panoramic view of the dynamic management of enterprise customers.Meanwhile, the ultra-large-scale customer environments, not just the data size isgrowing. Data types, data processing and analysis tool objects are changing. Thevalue of very low density data, the need for data integration, repeatedly taps thediscovery of new knowledge to identify potential value of data. Research enterprisecustomer relationship must adapt to the era of very large scale requirements ofcustomer data on the amount of data processing needs. The type of processingperformed on the processing speed of innovation and change in the way. In this paper,selected ultra-large-scale customer data environments massive scale customer dataanalysis to focus on objects at what Victor Meyer-Schonberg said, in massive dataenvironment, the best analysis method for improving a simple algorithm to make it suitable for knowledge of massive data mining algorithm selection to improve themain guiding ideology. K-means clustering algorithm because in massive dataenvironment, the impact of the effectiveness of the algorithm is computationallyefficient algorithm is large, so this paper selected time low complexity and highcomputational efficiency is improved object exists for the initial algorithm itself moresensitive and cluster centers identified in the mass data environment at low datadensity, high data sparsely adversely affect the algorithm will produce twoalgorithms to improve the algorithm, and the improved algorithm to conduct afeasibility analysis of the theoretical and applied aspects.