The Research Of The Factors And Prediction Model About Online Auction Price

Auction as a kind of effective market pricing mode has been widely recognizedby people. Since the1990s, with the rise of the Internet, ebay auction website as arepresentative of Internet auction began popular worldwide,now has formed a verypromising huge industry. Currently, the online auction has gradually become a majorform of online shopping.The transaction price of the auction is an important index tomeasure the auction results.The study of influencing factors on the price of theauction and online auction price forecast is the hot issue of the domestic and foreignscholars.By reference to the relevant literature and theory, based on analysis of thecharacteristics of the deployment of ebay auction online page and previous researchconclusion, this article presents a model of online auction price factors, created13proposition hypothesis, Factors into the description, auction property is set, the sellerof credit information, the seller’s level of service and market factors, and these factorsbe built a specific variable, and created the13propositions assumptions.We collectedCanon digital camera auction data and bid records on the ebay website. Throughmultiple linear regression analysis,Ultimately I retained nine influential factors, sevenfactors which network auction price, the end of the auction period, the number ofbidders has a positive impact, while the seller negative credit number and postagehave a negative impact on the network auction price,and draw the size of the impactof various factors.Next,I collected auction data sets on the ebay site, which isheterogeneity (Channel perfume and Nike sports shoes) and homogeneous goods(1922silver dollar coin of the dove of peace and the Canon EOS Rebel T3i/600Ddigital SLR camera).I combine with some of the the above conclusions about thefactors of online auction price, select the relative importance of the attributecharacteristics from the collected data set,build the online auction price predictionmodel by take advantage of the support vector machine classification. This forecastsystem can be implemented,used in the Internet price prediction, then provided timelyand useful information for the buyers and sellers in the process of bidding andauction.