EC Recommendation-System Based On Web Log Data Mining

Get into over 21 centuries, computer technique and computer network technique got to advance by leaps and bounds of development. Electronic commerce conduct and actions a kind of new business mode also develops immediately and quickly in the meantime.Web useage mining aims at picking up usage patterns from web usage througheffective web data mining. We can get valuable information we need such as usage patterns by analyzing the behavior, frequency and content of users' access on web. This information can contribute to the improvement of web design and, what's moreimportant; provide personalized services to users through the analysis andunderstanding of users' characterization.Electronic commerce because its cost is cheap, fast and is free from timespace restriction etc. the advantage gets universality and develops in the global scope, its scale also in the further extension. At the electronic commerce provides more and more choices for the customer of in the meantime, its structure also becomes more and more complicated, the development of electronic commerce faces thus a new problem-on the other hand, the customer isn't completely interested in numerous merchandise informations that the network provides, usually needing to pass many times to browse then can find out to satisfy an own demanding merchandise; On the other hand company's house can't understands personal need of customer completely, either, providing is a monotonous interface for customer, can't support stable customer to relate to. Lacking the characteristic service becomes check and supervision electronic commerce the key problem of the development. This will beg 1 and can analyze informations, such as customer hobby and behavior... etc., and the auto provide recommendation system of recommending the service toward the customer according to these informations, carrying out a characteristic network marketing.The Web usage mining be at thus of the background descend to combine with electronic commerce together. The Web data excavation discovers to sample also from the Web text file and the Web activity interested in of, latent useful mode with conceal of information. It puts together traditional technique and web knot of the data excavation, can Be producing result in many ways, is a new research direction that the data scoops out realm. The electronic commerce scooped out according to the Web data recommends system and can satisfy an electronic commerce future development the demand of the trend.This paper main to recommend system model according to the efficiency and the accuracy, and to system each function and they moderated a work to do a detailed description mutually;Went deep in to study electronic commerce recommendation the recommendation calculate way used by system, the point discussed currently use most for be in conjunction with to filter recommendation calculate way extensively;Designed on the foundation of above-mentioned research according to gather a type to be in conjunction with to filter to recommend system, and gathered a type of calculate way to carry on an improvement to the k-means;Make use of an actual website data finally to according to gather a type and be in conjunction with and filter recommend the system gathered a type of calculate way to carry on a realization, giving the system experiment a result, and to do andexplain as a result with evaluation.