Creek10171000

Slide1: 

Jeff Jonas, Chief Scientist and Founder Systems Research & Development Las Vegas, Nevada jjonas@srdnet.com SRD Copyright © 2003 Identity Fraud Solutions: Understanding People and Relationships and Knowing if They are Real

SRD Introduction: 

Founded in 1983. Headquartered in Las Vegas since 1992 Working with the gaming industry to better understand who they were doing business with since early 90’s Frequently featured in the media including: Wall Street Journal, Washington Post, Fortune Magazine, ComputerWorld, Discovery Channel, The Learning Channel, and MSNBC Funded by CIA’s In-Q-Tel to enhance technical capabilities Delivered identity warehouses supporting 4,200+ daily feeds, billions of rows, on multi-terabyte platforms Professional management team brought in to run the company SRD Introduction

Agenda: 

Agenda Cops and Robbers -- Las Vegas Style What is Entity Resolution™? Introduction to Non-Obvious Relationship Awareness™ (NORA™) Case Studies Challenging Problems Introduction to Anonymous Entity Resolution™ Questions and Answers

Targets & Threats: 

The Target Enormous amounts of cash change hands Outcome of transactions are probable but not certain High visitor volumes – perceived as “it’s easy to go unnoticed” The Threat Money laundering Cheaters and professional card counters Highly organized recruitment and training of new bad actors Credit and check fraud Insurance scams (e.g. slip and falls) Armed cage takeovers Executive kidnapping Targets & Threats

Slide5: 

Cops and Robbers Las Vegas Style

Highly Organized Criminals: 

Highly Organized Criminals

Slide7: 

What is Entity Resolution™?

Entity Resolution History: 

Entity Resolution History 1983: Technology originally developed to identify credit fraud for the credit bureau/collection industry 1994: Significantly enhanced for the gaming industry to protect them from unknowingly doing business with criminals and “insider threats” 1996: Created an identity warehouse for consumer brand giant. Entity resolutions being made against 4,200 data sources and 135,000,000 people. 2001: In-Q-Tel funding for certain technology enhancements September 2001: Used by a handful of US companies to generate leads of immediate investigatory importance to national law enforcement 2002: 2nd round In-Q-Tel funding for enhanced scalability 2003: Hit 400 Entity Resolutions/sec. against 340 million resolved entities

Entity Resolution: 

Entity Resolution Possible match on name only

Entity Resolution: 

Entity Resolution More data loaded

Entity Resolution: 

Entity Resolution No longer a possible match!

Entity Resolution: 

Entity Resolution Pat and Patricia determined to be same

Entity Resolution: 

Entity Resolution More data loaded

Entity Resolution: 

Entity Resolution Patrick is Pat

Entity Resolution: 

Entity Resolution Pat and Patricia are not the same!

Entity Resolution: 

Entity Resolution Juan and Miguel are not the same

Entity Resolution: 

Entity Resolution More data loaded

Entity Resolution: 

Entity Resolution Miguel J. is Juan Miguel

Entity Resolution: 

Entity Resolution Juan Tigar is Miguel J. Tigar!

Entity Resolution: 

Entity Resolution Mary is not deceased

Entity Resolution: 

Entity Resolution More data loaded

Entity Resolution: 

Entity Resolution We have learned Mary’s SSN and DOB

Entity Resolution: 

Entity Resolution Mary is deceased!

Slide24: 

With Entity Resolution More Data Makes More Context

Apparently Three Clusters: 

Apparently Three Clusters

More Data Enhances the Picture: 

More Data Enhances the Picture

Entity Resolved Creates Context: 

Entity Resolved Creates Context

Context Comparison: 

Context Comparison

Entity Resolution Principles: 

Entity Resolution Principles Context correcting – adjusts who is who as new data is discovered Sequence neutral – despite the order data received the results are the same Real-time 7x24x365 availability

Slide30: 

Non-Obvious Relationship Awareness™ (Determining Who’s Who and Who Knows Who)

Priority #1: Data Integrity: 

Priority #1: Data Integrity Address Hygiene NORA Database Name Standardization Data Quality/Enhancement Entity Resolution Load

Name Standardization: 

Name Standardization Mohamad, Mohammad Mohamed, Mohammed Mohammad Dick, Dickie, Ricardo, Rich, Richie, Rick, Rickey, Ricki, Rickie, Ricky, Rikki, Ritchie Richard

International Root Names: 

International Root Names African Arabic/Muslim Asian American American Indian Biblical Chinese Celtic Czech & Slovak Farsi French German Greek Hindi/Indian Hungarian Italian Irish Japanese Jewish/Hebrew Korean North Pacific Islander Pakistani Pashto Polish Polynesian Portuguese Romanian Russian Scandinavian Scottish Spanish Urdu Welsh

Address Hygiene: 

Address Hygiene 460 Oak Street Mill Valley, CA 94914 460 South Oak Ave Mill Valley, CA 94941 4737 Simeron Drive Easton, MA 02334 4737 Cimarron Drive Easton, MA 02334

Data Quality: 

Data Quality Formatting Standards Applied 7074121234 (707) 412-1234 557672061 557-67-2061 /SMITH&! SMITH Value Validation (213) 543- Rejected, incomplete (000) 000-0000 Rejected, invalid (800) 555-1212 Accepted, noted generic

Retain Historical Data: 

Retain Historical Data

Entity Resolution – Example Data Points: 

Entity Resolution – Example Data Points

Slide38: 

Good Guys The Watched Hotel reservations Hotel check-in’s Employees Vendors Victims OFAC SDN Nevada “Black Book” Gaming license revocations Griffin book Interpol Red List FBI Most Wanted

Perpetual Analytics (not data mining): 

Perpetual Analytics (not data mining)

Slide40: 

Case Studies

A Las Vegas Casino: 

A Las Vegas Casino DATA SOURCES: Employees Vendors Slot club and table games rated players In-house arrests/incidents database Industry published professional counters and cheaters DETECTED RELATIONSHIPS: 24 active players were known professional cheaters 23 players had relationships to prior arrests/incidents 12 employees were themselves the player 192 employees had possible vendor relationships 7 employees were the vendor

A Riverboat Casino: 

A Riverboat Casino DATA SOURCES: Employees Vendors Slot club and table games rated players In-house arrests/incidents database Industry published professional counters and cheaters DETECTED RELATIONSHIPS: 1 arrested roulette cheater lived with the dealer 1 “car-a-day” winner was sister of the promotion director 2 slot marketing hosts were providing cash back to their roommates

Slide43: 

Challenging Problems

Challenging Problems: 

Challenging Problems Identity packages False identities Stolen identities Data sharing policies Anonymity and personal privacy

Slide45: 

Introduction to Anonymous Entity Resolution™ (patents pending)

Slide46: 

Data Normalization One-way Hash Function Organization “A” Anonymous Entity Resolution (patents pending)

Anonymous Entity Resolution(patents pending): 

Anonymous Entity Resolution (patents pending) Data Normalization One-way Hash Function One-way Hash Function Organization “A” Organization“B” ALERTS Data Normalization Entity Resolution 96eea979df3929c77e1a44ce74600a83 f430b68d87d74191b9d62ff17a2e95ca 6f31b798c76ba550628c5f556a691301 e4e1c2e6db1616a385f1dcdc63a7df23

Slide48: 

Jeff Jonas, Chief Scientist and Founder Systems Research & Development Las Vegas, Nevada jjonas@srdnet.com SRD Copyright © 2003 Identity Fraud Solutions: Understanding People and Relationships and Knowing if They are Real QUESTIONS AND ANSWERS