Analysis And Implementation Of Shopping Online System Based On Hadoop

In today’s society, the development of science and technology more and mor e rapidly, with the application of technology also emerge in an endless stream, th e amount of data increases, the Internet has become one of the database giant ful ly deserve. This also represents the arrival of the era of big data. At present, the application of computer and broadband network technology based on the rapid em ergence of the update speed, technology, management is also very fast.At the moment, the "big data" to attract consumers to the use of the produ cts, many e-commerce companies can focus on how to achieve effective produ cts in big data information, the most typical of taobao. Sales data will directly affect the consumer’s behavior, people tend to buy products with high sales volum e, because the product is a lot of evaluation, it can help consumers tojudge whetherthe goods in line with their own. "Big data" will give consumers a high secur ity, in order to reduce the risk of online shopping. So, the electronic commerc e the pursuit of "big data" it is not surprising. Therefore, priority among priorit ies from big data is how to extract valuable data, but the Hadoop platform can effec tively solve this problem. The three advantage of big data technology:(1) the info rmation retrieval service function;(2) the accurate data analysis;(3) the elastic fastprocessing power.This paper mainly introduces the Hadoop based online shopping recommendthe working process of the system and the internal related research and analy sis of algorithms. Big data provides the 5 aspects of the value to the business e nterprise:(1) data of high transparency and availability;(2) effects of competitio n depends on the decision of verification;(3) are widely used and the influence o f real-time users;(4) a big data applications to change the way of management a nd replacement;(5) to build a business model based on large data.Main research contents:(1) The basic frame of Hadoop and the core components of Hadoop are st udied.(2) Analysis of the system structure and working principle of HDFS and M ap Reduce.(3) Based on Map Reduce framework, respectively on the naive Bayes algori thm, K-modes algorithm, ECLAT algorithm analysis. To investigate the principl e of large data stream running fast on Hadoop.(4) For these three algorithms, implementation of distributed algorithm, high efficiency and stable operation based on Hadoop platform, and of their compa rative analysis, to demonstrate its applicability to large data.