1 AT&T Proprietary Data Mining A pproach to Subscription Fraud Detection for AT&T Cards Hyunsook Lee, Summer Intern Risk and Revenue Modeling Group, AT&T.

Slides:



Advertisements
Similar presentations
Association Rules Spring Data Mining: What is it?  Two definitions:  The first one, classic and well-known, says that data mining is the nontrivial.
Advertisements

3rd Party Billing Provider Content Provider Mobile User Mobile Content Percentage of User Fee User Fee (monthly subscription or actual usage.
CS548 Spring 2015 Association Rule Mining Showcase Showcasing work by Ting, Pan, and Chou on "Finding Ideal Menu Items Assortments: An Empirical Application.
Chapter 4.
Data Mining Glen Shih CS157B Section 1 Dr. Sin-Min Lee April 4, 2006.
Chapter 9 Business Intelligence Systems
Giga-Mining Corinna Cortes and Daryl Pregibon AT&T Labs-Research Presented by: Kevin R. Gee 28 October 1999.
Data Mining By Archana Ketkar.
Roger S. Debreceny Shidler College of Business University of Hawai‘i at Mānoa Glen L. Gray College of Business & Economics California State University,
Effect of Bundling of New Telecommunications Service: A Customer Life-Cycle Perspective ITS 15th Biennial Conference Berlin, September 2004 Ann Skudlark,
U.S. Bank Payment Analytics Overview. Payment Fraud Trends 2 Reference: Association of Financial Professionals (AFP), 2011 Payments Fraud and Control.
Data Mining & Data Warehousing PresentedBy: Group 4 Kirk Bishop Joe Draskovich Amber Hottenroth Brandon Lee Stephen Pesavento.
TURKISH STATISTICAL INSTITUTE INFORMATION TECHNOLOGIES DEPARTMENT (Muscat, Oman) DATA MINING.
Enterprise systems infrastructure and architecture DT211 4
Governance, Risk, and Compliance Bill Greene Senior Industry Director.
Application of SAS®! Enterprise Miner™ in Credit Risk Analytics
『 Data Mining 』 By Jung, hae-sun. 1.Introduction 2.Definition 3.Data Mining Applications 4.Data Mining Tasks 5. Overview of the System 6. Data Mining.
MAKING THE BUSINESS BETTER Presented By Mohammed Dwikat DATA MINING Presented to Faculty of IT MIS Department An Najah National University.
10 Data Mining. What is Data Mining? “Data Mining is the process of selecting, exploring and modeling large amounts of data to uncover previously unknown.
M28- Categorical Analysis 1  Department of ISM, University of Alabama, Categorical Data.
Data Mining Dr. Chang Liu. What is Data Mining Data mining has been known by many different terms Data mining has been known by many different terms Knowledge.
CS490D: Introduction to Data Mining Prof. Chris Clifton April 14, 2004 Fraud and Misuse Detection.
Risk and Revenue Modeling Group AT&T Labs 12 - June AT&T Proprietary 1 Call detail: time of day, duration, origin, destination, charged number,
Kansas State University Department of Computing and Information Sciences CIS 830: Advanced Topics in Artificial Intelligence From Data Mining To Knowledge.
DATA MINING Team #1 Kristen Durst Mark Gillespie Banan Mandura University of DaytonMBA APR 09.
D ATA M INING A N O VERVIEW BY : J OSEPH C ASABONA Data Warehouse-->
Data and Process Modeling
Dynamic Black-Box Testing Part 2
Chapter 11 LEARNING FROM DATA. Chapter 11: Learning From Data Outline  The “Learning” Concept  Data Visualization  Neural Networks The Basics Supervised.
1 1 Slide Introduction to Data Mining and Business Intelligence.
TIA/EIA-124 (DMH) Background Network Reference Model Record Structure Data Types ASN.1.
Knowledge Discovery and Data Mining Evgueni Smirnov.
Copyright © 2013 by The McGraw-Hill Companies, Inc. All rights reserved. McGraw-Hill/Irwin Business Plug-In B18 Business Intelligence.
Payment workshop Identity, Security and Privacy Timothy Ng
Knowledge Discovery and Data Mining Evgueni Smirnov.
4. Secondary Data.
Marketing Research Marketing Information Systems.
Introduction to SQL Server Data Mining Nick Ward SQL Server & BI Product Specialist Microsoft Australia Nick Ward SQL Server & BI Product Specialist Microsoft.
Web Usage Mining for Semantic Web Personalization جینی شیره شعاعی زهرا.
Banking on Analytics Dr A S Ramasastri Director, IDRBT.
 Fundamentally, data mining is about processing data and identifying patterns and trends in that information so that you can decide or judge.  Data.
C6 Databases. 2 Traditional file environment Data Redundancy and Inconsistency: –Data redundancy: The presence of duplicate data in multiple data files.
MGT-491 QUANTITATIVE ANALYSIS AND RESEARCH FOR MANAGEMENT OSMAN BIN SAIF Session 16.
Data Mining – Intro. Course Overview Spatial Databases Temporal and Spatio-Temporal Databases Multimedia Databases Data Mining.
Guest Lecture Introduction to Data Mining Dr. Bhavani Thuraisingham September 17, 2010.
EXAM REVIEW MIS2502 Data Analytics. Exam What Tool to Use? Evaluating Decision Trees Association Rules Clustering.
3-1 Data Mining Kelby Lee. 3-2 Overview ¨ Transaction Database ¨ What is Data Mining ¨ Data Mining Primitives ¨ Data Mining Objectives ¨ Predictive Modeling.
Call detail: time of day, duration, origin, destination, charged number, … Scampweb CALLING PARTY CALLED PARTY 5E SWITCH CARD DATABASE CPP BILL SYSTEM.
DATA MINING By Cecilia Parng CS 157B.
1 Chapter 8: Introduction to Pattern Discovery 8.1 Introduction 8.2 Cluster Analysis 8.3 Market Basket Analysis (Self-Study)
MIS2502: Data Analytics Advanced Analytics - Introduction.
Some Final Material. GOOGLE FLU TRENDS Sore throat? Sniffles? Google it! Duh! During flu season, more people enter search queries concerning the flu.
WHAT IS DATA MINING?  The process of automatically extracting useful information from large amounts of data.  Uses traditional data analysis techniques.
Miloš Kotlar 2012/115 Single Layer Perceptron Linear Classifier.
Connect 3 Unit 2 Activity 2. . Connect 3: Grid Use dice to generate 2 digits. Use your digits to make 12 of the numbers on.
Introduction to Machine Learning Lecture 13 Introduction to Association Rules Albert Orriols i Puig Artificial.
DATA MINING It is a process of extracting interesting(non trivial, implicit, previously, unknown and useful ) information from any data repository. The.
Strategies for Metabolomic Data Analysis Dmitry Grapov, PhD.
Piecewise-defined Functions. In May 2003, Nicor Gas had the following rate schedule for natural gas usage in singe-family residences: – Monthly.
Data Mining – Introduction (contd…) Compiled By: Umair Yaqub Lecturer Govt. Murray College Sialkot.
Data Mining is the process of analyzing data and summarizing it into useful information Data Mining is usually used for extremely large sets of data It.
DATA MINING and VISUALIZATION Instructor: Dr. Matthew Iklé, Adams State University Remote Instructor: Dr. Hong Liu, Embry-Riddle Aeronautical University.
1  S. Matwin, 2002 Data Mining What is data mining? Motivating example Why now? Technological foundations Tasks Architectures and processes data warehouse,
Data Mining.
Data Mining, Machine Learning, Data Analysis, etc. scikit-learn
Data and Applications Security Introduction to Data Mining
Social Media Data Mining
Market Basket Analysis and Association Rules
Data Mining, Machine Learning, Data Analysis, etc. scikit-learn
Data Mining, Machine Learning, Data Analysis, etc. scikit-learn
Presentation transcript:

1 AT&T Proprietary Data Mining A pproach to Subscription Fraud Detection for AT&T Cards Hyunsook Lee, Summer Intern Risk and Revenue Modeling Group, AT&T Labs Supervised by Colin Goodall

2 AT&T Proprietary Objective: finding patterns in subscription fraud Contents a. Background b. Graphics c. Association Rules Discussion

3 AT&T Proprietary Data mining My definition : finding patterns or systematic relationships exploring data and TRANSFORMING them to indicators of interest Graphical Analysis Using DATA MINING TOOLS SAS Enterprise Miner

4 AT&T Proprietary Subscription Fraud Detection Analysis What is Fraud Subscription? Why the analysis is needed How to do it? a. Detecting subscription fraud from patterns of usage b. High Usage : Thresholding, but not only that… c. Other peculiar usage patterns : such as… d. Understanding calling cards e. Factors are possibly correlated Design and create new signatures graphics and association rules will help

5 AT&T Proprietary Data Sets & properties data sets: FASC, CARM FRAT : contains fraudulent info FPD : 1 st default payment data ( ):indicates business focus on FRAT data to find specific patterns of fraudsters FRATFPD #calls604141(8040) #cards7927(50)917 #accounts5053(34)551 Period358 days158 days

6 AT&T Proprietary Graphics.. # cards/(Paccount or BTN)

7

8 AT&T Proprietary Association Rules What are Association Rules ? a. customers’ item buying patterns b. support : P(A  B), confidence: P(A|B) How do we apply? a. analyze calls of each card and generate variables b. Variable generation based on idea from graphics and thresholding

9 AT&T Proprietary Variable generation & logics Possible characters of fraudulent cards a.Many international calls b.Zero Length calls, No Recorded calls c.Many calls d.Long duration, High rate e.Peculiar usage after certain period(such as 1 month) f.Satisfy $ based threshold, etc.

10 AT&T Proprietary NAMEDescriptionReference NonUSAAt least one calls made outside USAFPD, 3 rd INT> 4.3% international calls madeFPD, 3 rd Tint> 1.5% calls terminated to outside USAFPD, 3 rd OintAt least one calls originated to outside USAFPD, 3 rd NoRecAt least one calls recorded -1FPD, out ZeroLAll calls are zero lengthFPD, 3 rd BusH>54.5% calls made during business hoursFPD,3 rd LeisH>65.2% calls made during leisure hours(evening, weekend)FPD,3 rd NightH>8.9% calls made during night hoursFPD,3 rd WkEnd> 50% calls made during weekend WkDay> 50% calls made during weekdays TLNcard10 or 17 digits of card number TCcard6,7,8,9,16 digits of card number AT&T, Commercial, LEC : Billing Number Content …More variable can be generated…

11 AT&T Proprietary Results from by SAS Enterprise Miner

12 AT&T Proprietary Frequency of items

13 AT&T Proprietary Items generated by usage patterns, 60% confidence

14 AT&T Proprietary Future work Various approaches to generate Variables and Association Rules Classification methods are challenges: TREE, Random Forest…