The Design And Implementation Of Recommendation System Based On Taobao User Evaluation

With the rapid development of Taobao, Jingdong other e-commerce sites, more andmore consumers choose online shopping. Consumers can get their favorite products avariety of information on these e-commerce sites. Now, Taobao occupies the absoluteposition in e-commerce and has nearly500million registered users. Every day, shoppershave60million on e-commerce website, an average of48,000per minute merchandiseis sold. Taobao is committed to providing basic services platform, and other businessesinto the sales activities. No much qualification requirements of businesses, so Taobaoboth inside the famous brand’s flagship store, but also ordinary business, so how tochoose Taobao favorite products has become a very difficult problem. In the consumer’sonline shopping survey found that consumers can shop the most attention is thecommodity of the user evaluation, since commodity user rating reflects realistic outlookfor consumer goods, in order to determine overall quality of the goods. How toefficiently identify the consumer goods purchased comments emotional tendency forconsumers to buy goods have a very important significance.The main topic is the research and analysis of the commodity evaluationinformation, to find a new commodity recommendation system that allows users toquickly select favorite products. The main topics were analyzed from the followingaspects:To analyze for the current needs of consumers and Taobao shopping recommendordering their goods inconsistency problems: First, through the analysis of consumershopping and consumption habits of mental activity, and then get to the goods throughTaobao API evaluation data preprocessing, computing emotional value of the userevaluation information. Secondly, on the basis of Taobao existing comprehensivesorting algorithm improvements, depending on the weight ratio of each feature to beweighted, the formation of new merchandise comprehensive sorting algorithm. Finally,to build a platform interface display of goods, it is recommended a consumer favoriteproducts.Research on this subject abandoned Taobao based extant in popularity, sales, credit,price and other features are sorted and sorted based on these features integratedapproach avoids some businesses through PPC, speculation and other bad behavior get rankings front. At the same time considering the user evaluation emotional value ofgoods, sales, business reputation, and other factors, the author deeply analysis of therelationship between the factors, creatively presented based on each property’s overallmerchandise trade relations sorting algorithms. Theoretical analysis and experimentalresults show that research of this subject the user to find the product saves time, whileinhibiting the sale of counterfeit and shoddy products, superior in practical applications.