SAFARI Situational Awareness Framework for Risk Ranking Alberto - PowerPoint PPT Presentation

SAFARI Situational Awareness Framework for Risk Ranking Alberto Garcia-Robledo, Abel Sanchez, Rongsha Li, Juan-Carlos Murillo-Torres, John Williams and Sascha Boheme Massachusetts Institute of Technology MIT Geospatial Data Center z SAFARI: Situational Awareness Framework MIT Geospatial Data Center 1 for Risk Ranking

Challenges of Fraud Detection ● Few samples of fraud: i.e. unlabeled data (no supervised learning) ● Anomaly detection (unsupervised learning): high false positive rate ● Results may make no sense: what do the outliers actually mean? ● Reducing dimensions causes that data to lose its semantic meaning ● Finding a needle in a haystack How we can best exploit the unlabeled payment transaction douments considering the different types of information contained in them? SAFARI: Situational Awareness Framework MIT Geospatial Data Center 2 for Risk Ranking

Situational Awareness for Fraud Detection SAFARI introduces the concept of Situational Awareness to enable the detection of fraud on large volumes of payments where ground truth is not available, by integrating different perspectives of financial data . SAFARI: Situational Awareness Framework MIT Geospatial Data Center 3 for Risk Ranking

SAFARI: 1 and 2-Level SA Approach SA Risk Managment and Prediction Situational Awareness Level 3: Projection Predictive Analytics Level 2: Comprehension RFNet modelling Visual Analytics and Web dashboard and Data Integration visualizations Level 1: Perception RF raising at Data Collection and Anomaly Detection different perspectives of data SAFARI: Situational Awareness Framework MIT Geospatial Data Center 4 for Risk Ranking

SAFARI: Workflow PRC Payment Docs. GXM Payment Docs. GAX Payment Docs. ... SAFARI: Situational Awareness Framework MIT Geospatial Data Center 5 for Risk Ranking

SAFARI: Data Ingestion and Enrichment Enricher SAFARI: Situational Awareness Framework MIT Geospatial Data Center 6 for Risk Ranking

SAFARI: Anomaly Detection N-way Exact Matching N-way N-way String Fuzzy Matching N-way N-way String Phonetic Matching N-way N-way Address Geolocation Matching Rule Expression Matching N-way N-way N-way N-way SAFARI: Situational Awareness Framework Accenture | MIT Alliance in Business Analytics 7 for Risk Ranking SAFARI: Situational Awareness Framework MIT Geospatial Data Center 7 for Risk Ranking

SAFARI: RFs and RFNet Integration RFNet 0 SAFARI: Situational Awareness Framework MIT Geospatial Data Center 8 for Risk Ranking

SAFARI: RFNet Scenarios RFNet 3 G M K RFNet 2 F J I RFNet 4 H L N RFNet 5 RFNet 1 E P Q O C A D R B SAFARI: Situational Awareness Framework MIT Geospatial Data Center 9 for Risk Ranking

RFNet Pair-Wise Matching and BBN Ranking P( a = T) = 1 FuzzyA A NameFuzzyMatch P( e = T) = 1 a NameMatch P( b = T) = 1 e A PhonA NamePhonMatch b P( r = T) = 0.95 VendorMatch r = 0.95 r P( c = T) = 1 GeoA AddressGeoMatch P(f = T) = 0.8 c AddressMatch B P( d = T) = 0 f AddressNwayMatch B d SAFARI: Situational Awareness Framework MIT Geospatial Data Center 10 for Risk Ranking

SAFARI: Risk Score Propagation D D 0.91 0.91 0.91 C 0.84 C 0.84 E 0.91 E 0.65 0.84 0.65 B 0.65 F 0.23 B 0.65 0.95 F 0.95 0.23 0.95 0.23 G 0.23 0.23 0.65 G A A 0.95 RFNet score: 0.95 SAFARI: Situational Awareness Framework MIT Geospatial Data Center 11 for Risk Ranking

SAFARI: Finding the Needle ... Risk: Low G M K Risk: Low F J I Risk: Low H L N Risk: Low Risk: High P E Q O C A D R B SAFARI: Situational Awareness Framework MIT Geospatial Data Center 12 for Risk Ranking

SAFARI: Finding the Needle ... Risk: Low G M K Risk: Low F J I Risk: Low H L N Risk: Low Risk: High P E Q O C A D R B SAFARI: Situational Awareness Framework MIT Geospatial Data Center 13 for Risk Ranking

SAFARI: Web-Based Visual Analytics SAFARI: Situational Awareness Framework MIT Geospatial Data Center 14 for Risk Ranking

Conclusions • Analysis integration . Combine different analysis techniques for processing large amounts of payment documents. • Big data analysis . Help SMEs to make sense of a large amount of RFs spread across data. • Focus . Help SMEs to focus on the most suspicious payments by exploiting modern high-performance multi-core computers and visualization techniques. Integration False positive Ranking Novelty: minimization Visualization SAFARI: Situational Awareness Framework MIT Geospatial Data Center 15 for Risk Ranking