Prediction Models on Internet Auctions Prediction Models on Internet Auctions Prediction Models on Internet Auctions Prediction Models on Internet Auctions

AbstractMostinternetauctionsitesopentousersthebidhistorywiththeascendingorderofbidamounts.WhereaseBay.compresentssecondbidprices,auction.co.krprovideshighestbidprices.Inthispaper,thebidhistoryisarrangedaccordingtothepassageoftime,whichcanhelptounderstandthesituationsandtrendsofbidprices,especiallyformultipleauctions.Thismanipulateddatacanbevisualizedbyusingprofileplots.Thesuccessfulbidpricescouldbeestimatedbasedonsomepredictionmodelswithappropriatepriorinformations.Bothsellersandbidderscanbeprovidedusefulinformationswiththesestatisticalanalyses,andthenfaironlineauctionsinKoreawillgrowactivelyandrapidly.Keywords:Firstbidamount,Multipleauction,Profileplot,Regression1.11..1. 서론 각국마다많은인터넷경매사이트가있으며 미국에서는인터넷경매사이트인이,베이 를한국에서는 주옥션 을많이이용하면서물건들을경(eBay.com) ( ) (auction.co.kr)매한다 한국의인터넷경매에대하여많은논문이발표되었지만대부분의논문은경.매시스템설계와구현그리고여러종류의경매모형등에초점을둔정보공학과컴퓨터공학또는경영학과산업공학관점의논문이대부분이다이종회외 인 박( 3 (1999),용진외 인 이용준외 인 김광수와김인철 이근왕외 인 이2 (2000), 3 (2000), (2001), 3 (2002),근호와김태윤 유양우외 인 이영균과공은배 조경산과원유석(2001), 4 (2002), (2002),(2002), (2003), 2 (2003), (2005),성순화와공은배 이연수외 인 고민정과이용규 이기환과임동균 등참조 와 는미국의경매회사인이베이의자료에(2005) ).Shmuelli Jank(2005)대해통계분석하는여러방법을제시하였는데 경매한기록을종합하여전반적인추,세및흐름을파악할수있는방법인 과경매달력 등을STAT-zoom (Auctioncalendar)제안하였으며특히 다양한윤곽그림 을개발하였다 윤곽그림에는시간의, (profileplot) .