Research And Application Of Web Mining In E-commerce

With the rapid development of the mobile network, the tide of shopping online is on the upgrade. The development of E-commerce plays an very important role in domestic economy. And the processes of interactions between users and web services produce large amounts of data which seems very messy but contains a huge commercial value. The e-commerce web urgently need to mine their own data and build a association data, for winning a favorable advantage in the modern. The technology of web data mining has become a hot in higher universities and enterprises. According to actual needs, compiling the spider to crawl data in websites of Taobao.Firstly, web data mining applies in the websites of e-commerce especially in the system of personalized recommendation. These are different types of the system depending on the different algorithm these recommendation systems have pros and cons. Currently the association rule algorithm based on item-item and user-item is widely used in e-commerce websites. Secondly, the fuzzy clustering algorithm is applied in dividing customer groups. Through processing customers’ transaction data and compiling the code to fulfill the fuzzy c-means algorithm and classified the raw data. According the clustering results, elements in the same cluster have higher similarity, while the elements in the different cluster have lower similarity. Identifying the properties of different clusters depending on the outcome. And enterprises learn of customers’ preference in color and their consumption interval and their loyalty to the website and so on. Achieving a better clustering effect though fuzzy c-means cluster. Meanwhile the supervisor can make different strategy relying on different clusters. Finally, the technique of data extraction from web pages, security policy and the mainly modules of web crawler including to data storage modules, spiders module and deployment of real-time minor modules. Run the spider to crawl data from pages in 30 days. Compiling those data with the power law and making predictive analysis with times series. The time-series data is mapped to complex networks in visualization algorithm. Research the distinguishing feature of sales data in time series and data’ fluctuation. Focus on consumers’ sensitivity to price and sales campaign.