English  |  正體中文  |  简体中文  |  Items with full text/Total items : 57322/90918 (63%)
Visitors : 13031935      Online Users : 205
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library & TKU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version
    Please use this identifier to cite or link to this item: http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/102396

    Title: 一套有效率的複合式線上拍賣詐騙偵測系統
    Other Titles: An effective composite fraud detection system for on-line auctions
    Authors: 林敬堯;Lin, Ching-Yao
    Contributors: 淡江大學資訊管理學系碩士班
    Keywords: 詐騙偵測;分類樹;線上拍賣;電子商務;Fraud Detection;Decision tree;Online Auction;Electronic Commerce
    Date: 2014
    Issue Date: 2015-05-04 09:54:44 (UTC+8)
    Abstract: 近年來,線上拍賣的蓬勃發展有目共睹,交易量屢創新高。但在此同時,詐騙者開始進入此一便利的交易平台,利用網路的隱蔽性大肆進行不法活動。詐騙手法不斷推陳出新,甚至配合早被遺忘的陳年手法,不斷循環運用,讓有經驗的交易者也難以識破。為了識破偽裝完善的狡猾詐騙者,本研究提出了二套不同偵測方法: 分群匹配塑模法與名聲因子分類組合法,以不同的方式組合多種不同的偵測模型,以提昇詐騙偵測的準確率。在分群匹配塑模法中,訓練資料集事先以群集演算進行分群,再根據待測帳號的特性,即時建構出最合適的分類模型。名聲因子分類組合法則根據常用的拍賣者名聲模型,將分類模型分為年份、評價分數與交易類別等三種,並以協力的方式來過濾詐騙者,透過投票或權重組合來判別可疑者的身分。為了驗證方法的有效性,我們使用YAHOO!奇摩的真實交易資料進行實驗。結果顯示,本研究提出的方法能有效提升詐騙者偵測的精度,並保持優良的總體偵測成功率。
    In recent years, the rapid growth of online auctions were seen by everyone. Trading volume hit record highs. But in the meantime, began to enter a fraudster convenient trading platform, using a network of hidden wantonly engaged in illegal activities. Scams constantly, even with long-forgotten vintage approach, continuous cycle of use , so that experienced traders also difficult to see through . Although many online auction platform precautions , but most of its design -oriented seller by the buyer cheated in order to begin to remedy the cause scammers have nothing to fear , rampant . In order to see through the disguise perfect cunning scammers , this study combined in different ways in many different classification tree , more accurate fraud detection. Type Total Year , evaluation scores and transaction types , three classification tree to generalize traders at different times of the various characteristics of different types of transactions . The classification tree is not used alone , but in a third way to filter fraudster . Combination of various ways depending on the type of classification trees to re- vote or the right to determine the combination of suspicious persons identity. In order to verify the effectiveness of the method , we use YAHOO! Kimo ''s real transaction data to validate the experimental results show that our proposed method can effectively improve the accuracy of detection scammers and maintaining excellent overall detection success rate.
    Appears in Collections:[Graduate Institute & Department of Information Management] Thesis

    Files in This Item:

    File SizeFormat

    All items in 機構典藏 are protected by copyright, with all rights reserved.

    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library & TKU Library IR teams. Copyright ©   - Feedback