Business AnalyticsIBM Software IBM SPSS Modeler

Entity AnalyticsGreatly increase the accuracy of your models quickly and easily

Entity Analytics

2

Contents

2 Entity Analytics — because context matters

3 About IBM SPSS Modeler Premium

3 Entity Analytics functionality in IBM SPSS Modeler Premium

5 Hypothetical bank lending scenario using Entity Analytics

7 Real-time Entity Analytics

7 Summary

Entity Analytics – because context mattersAnalysts routinely face steep challenges as they attempt to integrate diverse enterprise-wide data – especially when this data contains natural variability (for instance, Bob versus Robert), unintentional errors (like a transposed month and day in a date of birth) and professionally fabricated lies (such as a fake identity).

Entity Analytics allows analysts to overcome some of the toughest data preparation challenges with unprecedented ease. Using Entity Analytics, analysts are able to generate higher quality analytic models resulting in better business outcomes – whether the goal is detecting and preempting risk or recognizing and responding to opportunity.

One of the more important data preparation activities involves recognizing when multiple references to the same entity are the same entity (within the same and across data sources). For example, it is essential to understand the difference between three transactions carried out by three people versus one person who carried out all three transactions.

After determining when entities are the same (resolved), even deeper understanding is achieved by recognizing when these resolved entities are related to each other (such as sharing a home address). Going far beyond simplistic match/merge technologies of the past, today’s Entity Analytics delivers something new: true “context accumulation.” Context accumulation is the incremental process of relating new data to previous data and remembering these relationships. In other words, you can understand something better by taking into account the things around it.

For example, a stand-alone puzzle piece can be difficult to evaluate for importance when staring at the piece by itself. However, by first comparing the puzzle piece to the whole puzzle to see how it relates to the previously seen puzzle pieces, the observer can better understand the bigger picture and make a better prediction.

Figure 1: Entity Analytics accumulates context over diverse data.

2

Business Analytics

3

About IBM SPSS Modeler PremiumIBM® SPSS® Modeler Premium is a high-performance predictive and text analytics workbench that is designed to help you gain unprecedented insight from your data. It provides a broad set of analytic capabilities including:

• Visualization and exploration of data• Data manipulation• Cleaning and transformation of data• Creation and evaluation of predictive models• Deployment of results in the form of production (run-time)

models or scores

Entity Analytics functionality in IBM SPSS Modeler PremiumSPSS Modeler Premium contains Entity Analytics capabilities that enable analysts to quickly associate identity, behavior and action data with their respective entities in real-time or batch, with extraordinary ease.

These entity analytics capabilities in SPSS Modeler Premium represent a breakthrough technology – the first of its kind commercially available. Even better, these capabilities are easy to use, allowing you to start taking advantage of these features right away.

These entity analytics capabilities in SPSS Modeler Premium represent a breakthrough technology – the first of its kind commercially available.

Analysts have historically been spending up to 80 per cent of their time preparing and cleaning data for analysis. Using Entity Analytics, users can now build much more accurate models, based on cleaner data in a shorter time frame. Users of Entity Analytics gain these distinct advantages:

• Moreaccuratepicture. The more identifiers that accumulate for an entity, the more accurate the Entity Analytics technology becomes.

• Bettermodels. Information in context (understanding how the data relates) delivers higher quality models.

• Betteroutcomes. Higher quality models applied to the context enhanced transactions produce better decisions (for instance, risk scores).

Entity Analytics

4

As an illustration, it is a common regulatory practice to require banks to report all cash transactions over $10,000. In order to comply, banks must be able to understand the difference between ten unrelated $5,000 cash transactions versus one person transacting $50,000. If a bank cannot accurately quantify the cumulative (historical) transactions for that individual, it will be unable to determine whether the $10,000 threshold has been crossed.

Entity Analytics provides an exceptionally easy means (using context accumulation) to associate all of the transactions to a common entity, despite the lack of a common key (in other words, the accounts do not share a tax ID number). As a result, when the transactions are in context, the scoring models operate on the $50,000 number – not a seemingly unrelated number of $5,000 transactions.

Figure 2: An example of an SPSS Modeler Premium stream using Entity Analytics.

4

Business Analytics

5

In SPSS Modeler Premium, Entity Analytics can be in three different ways:

1. The EntityAnalyticsExportNode performs the context accumulation – determining whether two different entities (such as, individuals, corporations or vehicles) are the same, despite the fact they have been recorded separately and to a degree, differently. If entities are determined to be the same, the entity’s identifiers (such as name, address or phone) and the measurements (such as average balance or credit limit) are accumulated for that entity. This node automatically applies sophisticated fuzzy matching techniques. For example, it takes advantage of an internal library based on 800+ million people names to deliver a world-class culturally-aware name comparison. As entities are resolved, understanding about each entity improves. This node will frequently be used to integrate historical data from the past with incremental data going forward.

2. The EntityAnalyticsSourceNode allows the analyst to access the “in context” information – a registry of what entities exist and which entities are deemed the same. This node will frequently be used to analyze historical information (in context) when building new models.

3. The StreamingEntityAnalyticsNode is used to apply new records (in batch or real-time) to the historical information (current to the second). It instantly recognizes when entities are the same or related. This is roughly akin to being given a piece of data (a puzzle piece) and asking what its resume (the other related puzzle pieces) might look like. This node will frequently be used to evaluate a current transaction (for instance, to assign a risk score) more accurately, thanks to the enhanced context.

Hypothetical bank lending scenario using Entity AnalyticsTo see how Entity Analytics really works, the following is a hypothetical example involving a typical bank process: loans to customers.

Predictive Analytics can be used to help a bank determine which customers are likely to pay back their loans versus defaulting. Using historical customer data (for instance, income, debt or previous defaults), past lending outcomes (such as, credit limit, average payment amount or delinquencies) and other frequently-used data points, models are developed to determine the likelihood of an individual paying back a loan. Figure 3 is an example of some historic lending data.

CustomerNumber

Income Credit Debt

Other Debt

Debt To Income

Prev Default

Pending

102 8000 5359 2009 92.1 Y N

343 9000 6000 3000 100 Y N

642 31000 1362 4001 17.3 N Y

Figure 3: Historic lending data.

Entity Analytics

6

Looking closely at Figure 3, imagine the difference between the three data points about three different customers versus these three data points about the same customer. Would you consider customer #642’s pending credit application a credit risk if you knew with some confidence that he has already defaulted twice in the past (assuming customer #642 is in fact, the same person as #102 and #343)?

If customers used their true names, addresses and identifiers consistently and provided all details comprehensively and unambiguously, determining that these are the same customer would be trivial. Unfortunately, because of unintentional data quality issues and periodic criminal intent, determining that this is the same customer is easier said than done. Fortunately, Entity Analytics allows users to quickly and easily perform context accumulation – detecting exactly this sort of situation (and more). In Figure 4, we see that entities #102, #343 and #642 share sufficient identifiers to make a very strong claim this is the same customer.

Figure 4: Common attributes across diverse records are used to construct context.

Business Analytics

6 7

Using these collected facts together in the “Resolved Entity” column brings to light essential context to help score this pending loan application for entity #642 properly. This is the true value of Entity Analytics: more accurate decisions, faster.

This is the true value of Entity Analytics: more accurate decisions, faster.

Real-time Entity AnalyticsUsing Entity Analytics, SPSS Modeler Premium allows companies to analyze transactions in real-time to make optimal decisions in context. By using all the “big picture” information, the models are able to predict outcomes more accurately for instant decision making, such as real-time fraud detection.

Imagine a fraud investigator who has just stumbled upon a new address related to an ongoing internal criminal investigation. Would it not be amazing if seconds later Entity Analytics instantly alerted this investigator that an employee in their very own credit department has the same address? Context accumulation over enterprise data like customers, employees and investigations delivers this kind of extraordinary “insider threat” insight and so much more.

SummaryThe Entity Analytics feature in IBM SPSS Modeler Premium allows analysts to pull diverse enterprise data together into context. Organizations can then use this information in context

to improve model quality, make better decision and ultimately achieve greater success – whether the objective is mitigating risk or recognizing opportunity.

An organization that can make sense of what it knows and do something about it faster than the competition is more competitive. With this exciting new technology, organizations of all sizes can gain this competitive edge today.

About IBM Business AnalyticsIBM Business Analytics software delivers actionable insights decision-makers need to achieve better business performance. IBM offers a comprehensive, unified portfolio of business intelligence, predictive and advanced analytics, financial performance and strategy management, governance, risk and compliance and analytic applications.

With IBM software, companies can spot trends, patterns and anomalies, compare “what if” scenarios, predict potential threats and opportunities, identify and manage key business risks and plan, budget and forecast resources. With these deep analytic capabilities our customers around the world can better understand, anticipate and shape business outcomes.

For more informationFor further information, visit ibm.com/business-analytics.

Request a callTo request a call or to ask a question, go to ibm.com/business-analytics/contactus. An IBM representative will respond to your inquiry within two business days.

YTW03247-USEN-00

© Copyright IBM Corporation 2012

IBM Corporation Software Group Route 100 Somers, NY 10589

Produced in the United States of America July 2012

IBM, the IBM logo, ibm.com, Cognos and SPSS are trademarks of International Business Machines Corp., registered in many jurisdictions worldwide. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on the Web at “Copyright and trademark information” at ibm.com/legal/copytrade.shtml

This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates.

THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided.

Please Recycle