Data analysis for game fraud detection
- 1. Data Analysis for Game Fraud Detection NCSOFT Data Platform div. Lee Eunjo
- 2. What is fraud detection? 豢豌: http://www.canadianunderwriter.ca/news/fraud-detection-software-good-investment-if-claims-staff-know-how-to-use-it/1002126531/
- 3. るジ 一 覿覯朱 豬 豺企 螻譬 覲危 蠍 豢豌: http://www.healthcareinfosecurity.com/oig-ehr-fraud-detection-inadequate-a-6358
- 4. 200 豌 殊 蟇磯 9% (蟇磯 觜 襷る 2覦 煙) (Online Fraud Report, CyberSource, 2012)
- 5. Fraud detection : Fraud襯 讌 譯 覿 蠍牛 旧 蟆一 豢豌: http://www.the-hard-life-of-a-check.com/fraud-detection-in-distributed-capture/
- 6. Fraud detection method if-then rules predictive modeling abnormality detection network analysis 豢豌: http://www.propertycasualty360.com/2012/12/06/technology-claims-segmentation-enhance-fraud-detec
- 7. IF-THEN Rules (expert rules) 企 覿 覓瑚 蟆渚 讌 蠏豺 蟆一 企 覓語襦襷 螳? 豺企襯 り覃 蟆一螳? 企 譯殊 螳 襷 螳? 蟆一 螳 覯曙瑚? 5覿 企伎 2蟇 伎 蟇磯螳 覦螳? 讌 蠏豺 螻 蟆郁骸 伎 覦覃 襦 蠍 讌 覿螳 豢豌: http://www.todayandtomorrow.net/2009/03/12/if-then/
- 8. Predictive modeling 糾 / 一危磯企 蠍磯 伎 讌 Decision Trees Random Forest Logistic Regression Neural Network Training Set Algorithm ModelTransaction Data 讌 蟲豢 企れ 豢豌: http://www.propertycasualty360.com/2011/06/20/innovations-in-claim-fraud-detection
- 9. Abnormality detection るジ 轟 るジ 螻 襷れ るジ Outlier 讌 - 20 螳 蟇一 煙 襴讌 蟆 危 蟲襷 覦 襦朱 企れ 豢豌: http://www.pskcpa.com/fraud-prevention/church-fraud-detection-con-man-signs/豢豌: http://www.accertify.com/Solutions/Fraud-Management/card-not-present-fraud/
- 10. Network analysis Fraudster 螳 蟯螻 ろ語 覿 朱碁り骸 るジ ろ語 轟 讌 覿 蠍磯 覩 煙
- 11. 讌襷 れ 蠍 襦 讌 豌 殊 蟇磯 80% ろ(False positive) 200 蟇磯螳 ろ 蟆一 蟇磯 襭 豢豌: http://www.slideshare.net/Datameer/online-fraud-detection-using-big-data-analytics-webinar-32814816 豢豌: http://falsepositivecomic.com/2011/09/10/season-1-title-page/
- 12. Game Fraud
- 13. 螻 るジ 蟆 螻 伎 覲伎 蟆 危/一 覈磯 豌覿 蟆 企薫讌 蟆 覯蠏碁ゼ 覈磯 蟆 危 覓危 覲旧, 覲伎 覈 覓危
- 14. 襦蠏碁(ろ) 覓伎 る 襴も
- 15. 譴 る 蠍壱 蟆
- 16. 蠏碁る Game Fraud Detection 企至 螳? 豢豌: http://www.chasepaymentech.com/fraud_online_retail.html
- 17. 蠍磯蓋朱 螳 襷 Expert Rules 螳螻 讌 覲伎 企れ (企轟螳 伎覃癌?) 讌 蠏豺 覲伎 (螻糾覃 ) 豢豌: http://www.virginiatitlecenter.com/alta-best-practices/
- 19. Self similarity algorithm 螳 貂襴 覲 蠍 覦覲旧 蟆渚 ろ 貂襴磯 覩碁Μ れ 襯 覦覲牛 豺螳 貂襴磯 讌 蠍 豢豌: http://mark.rehorst.com/Bug_Photos/index.html
- 20. 貂襴 覲 覦 襦蠏碁ゼ 覯″磯 覲 -> 螳 覯″磯れ Cosine similarity 螻 Normal User BOT Self similarity algorithm (0, 1, 1, 3) (2, 1, 1, 2) (0, 1, 2, 1) (0, 0, 0, 4) (0, 1, 0, 0) = = | | = = () = ( ) = A type 襦蠏 覦 螳 B type 襦蠏 覦 螳 E (1,1) Vt (2,1) cos(慮)
- 21. 貂襴 覲 Cosine similarity 譴 ｸ谿 螻 蠍 螳朱 覲 Self similarity algorithm
- 22. 讌 伎 蠍一 螳 BOT 襯襦 Self similarity algorithm Character ID T1 T2 T3 豬蟆郁骸 686042 0 0 0 0 854209 1 1 1 3 1032131 0 0 0 0 1049483 1 1 1 3 1340479 0 0 0 0 1352850 0 0 0 0 1771815 1 1 1 3 1832497 0 0 0 0 1884884 1 1 1 3 2130576 1 1 1 3 2445903 1 0 0 1 貂襴 覲 ろ 襯
- 23. 貂襴 覲 蠍 覲 豢 覿 Self similarity algorithm ろ襯 觜 朱 伎 螻 讌 ろ 朱 ろ 蟲覿
- 24. Self similarity algorithm 覲 讌 覦讌 覦 螻手碓 ろ 襯螻殊 蟯 螻 覲 豢 (EWMA*) 伎 覯襯 伎 覲 覦 覈 轟 讌 蟲豢 *EWMA: Exponentially Weighted Moving Average
- 26. Banker Character Detection 豢豌: http://www.rollingstone.com/politics/blogs/taibblog/everything-is-rigged-vol-9-713-this-time-its-currencies-20130613
- 27. Banker Character Detection BOT BOT BOT Banker Banker Banker Normal User Normal User Normal User Monetization Trading Trading Item Shop Item Selling
- 28. Banker Character Detection BOT BOT BOT Banker Banker Banker Normal User Normal User Normal User Monetization Trading Trading Item Shop Item Selling 蠍一ヾ 讌 ろ 蟆豢 覿螳
- 29. Banker Character Detection BOT BOT BOT Banker Banker Banker Normal User Normal User Normal User Monetization Trading Trading Item Shop Item Selling Graph Clustering
- 30. Banker Character Detection 貂襴磯り 蟇磯 ろ語 蟲 蠏碁 企ろ磯 ろ螳 襷 企ろ() 蟲覿 Banker 貂襴 讌
- 33. Banker Character Detection Banker 貂襴 襯 牛 襯 螻 レ れ 蟆曙 蟆
- 34. Fraud detection 覈 螳讌 螳 豢豌: http://www.propertycasualty360.com/2013/10/01/five-steps-to-a-better-fraud-detection-program
- 35. 螳 Fraud detection 企 謂仰
- 36. False Positive れ 朱 襯 ろ手 覃? 襦 蟆 豺蟲蟆 譬 危 譴 襯 麹覃? 豢豌: http://www.threatmetrix.com/marketing-fraud-detection-friction-or-faction/
- 37. False Positive れ 朱 襯 ろ手 覃? 襦 蟆 豺蟲蟆 譬 危 譴 襯 麹覃?
- 38. 覦覃 豢豌 覃 譬螻 覃 襷螻
- 39. Target 炎概 襦 蠏朱 ろ 觜 朱 讌 れ
- 40. Fraudster 讌 伎 狩蠍 覲 Never-ending Change
- 41. Never-ending Change 麹 讌 螻手 伎襦 蠏轟朱 覲 ( 襦 れ伎 襷)
- 42. 谿瑚 一危 覦 襦 覿譟 一危 蟲豢 企れ 襷 企れ 襭 螻旧螳 蟇一 豢豌: http://bennettandbelfort.com/blog/?p=482
- 43. 讌 蟲豢 襷れ 譴 Fraud 一危一 一危 觜 襷 讌 一危 豢豌: http://www.nkkhoo.com/2012/12/09/garbage-in-garbage-out/
- 44. 蟯 螻 螳 覲 蟲襯 ルる 企蟆 覦 蟲襯謂
- 45. 螳 一危 螻糾襯 牛 覦 蠍一 覦 企譟朱癌 豢豌: http://en.wikipedia.org/wiki/Open_data
- 46. 螳矧