Exploring Address Screening Functionality 

Address screening compares business partners or transactions with lists of certain parties to fulfill legal obligations.

In most countries or regions, governments legally require companies to compare their business partners to lists of terrorists, sanctioned parties, politically exposed persons, and so on, to enforce security measures and political bans in daily trading. The software enables you to screen any business partner in your business documents against such lists.

To perform address screening, you need the following:

• One or more address screening lists. You can obtain these lists of persons and organizations of interest from third-party data providers and upload them into your system, or you can create your own lists manually.
• Customizing that groups the content in the lists into list type groups. This enables you to use parts of the same loaded lists in different scenarios.
• An address view that retrieves the names and addresses from your data. 
• One or more detection methods for address screening. You can create detection methods for address screening using the Manage Detection Methods app.

Optionally, you can improve your screening quality by using:

• Additional screening attributes like date of birth or gender by implementing address filters
• Excluded terms to leave out certain terms; for example, to ignore common abbreviations such as "Corp." or "Ltd.".
• Term mappings to treat terms in the same way; for example, to treat the abbreviation "Corp." and its long form "Corporation" equally.

BIS Address Screening integrates with any kind of List Type from any List Provider.

List Providers typically provide a broad range of different list types like PEP, Sanction Lists, Adverse-Media, State-Owned-Companies etc.

SAP BIS solution uses concept of list types and list type groups. List types are mapped to the provider-defined list groups and list type groups are used in detection strategies.

This allow flexible assignment of lists to detection strategies (to allow for example one strategy to be used for PEP-lists and another one for sanction lists).

The following figure illustrates this concept.

You can visualize the mapping between lists, types and strategies using the Assign Address List Types app.

BIS can now be configured with a Detection Strategy to make use of the list-type specific entities and perform Address Screening with the Standard screening framework of BIS.

You can define various screening strategies for various risk levels, business scenarios and screening lists.

Customers typically do have low-risk list types and high-risk list types. With the flexible Strategy configuration, you can assign different Risk Scores to different kind of list types.

• For e-commerce system in Japan, screen against OFAC and METI list with low thresholds
• For other e-commerce systems, screen against all sanctioned lists with low thresholds
• For all the systems, screen against PEP lists with very high thresholds

As mentioned, there are two actors in the process: Business partners and Screening list entities.

Business partners come from various systems. And depending on the underlying transaction they pose different levels of risks.

For example, risk of a person downloading free manual from your website is completely different from the risk of the same person downloading encrypted software or the same person ordering a product from your website.

In SAP BIS solution you can define detection strategies for various risk levels, business scenarios and screening lists.

You use the selection parameters (business attributes) to identify business partner or transactions to be screened by the strategy and list type groups to identify specific list.

Screening parameters are available in the Address Screening Methods out-of-the-box and allows business user to configure the screening logic and optimize screening results.

The following parameters are available in the SAP BIS solution:

Exactness

The exactness (technical term FUZZINESS) is a percentage between 0 and 100 that specifies how exact two words must match to be considered as equal. Use this parameter to make the screening tolerant towards typos, name variations, accents, umlauts, and so on. The lower the value, the more tolerant the system is. For example, an exactness of 70% will match Jane to June, while 90% will not.

Minimum Score

The minimum score (technical term MINIMATCH) is a number between 0 and 100 that specifies how precise two names or two addresses must match to be considered as a hit. In contrast to exactness, this parameter affects the names and addresses as a whole, not the single words therein. The lower the value, the more two names or addresses can differ and still be a hit. For example, a minimum score of 50 will produce Tomas Meyer as a hit for Thomas Mayer, one of 100 will not. Use this parameter to cut off the long tail of low-quality hits produced by low exactness parameter values.

Percentage of Matching Words

The percentage of matching words (technical term ANDTHRESHOLD) is a value between 0 and 100 that specifies what percentage of the words making up an entire name or address must match. The lower the value, the more reactive the screening is towards single words. For example, percentage of matching words of 66 will produce John Richard Adams as a hit for Michael John Adams, one of 80 will not.

Include Term Mappings

• Include term mappings (technical name ALIASES) is a yes-no value that specifies whether the screening shall enrich the search string with additional terms from the term mappings. Term mappings can be used to prevent that the screening is bypassed by common abbreviations and misspellings. For example, the value Y makes the screening match Main Street to Main St. if the term mapping maps St. to Street.
• With the app Manage Term Mappings you can create lists of terms to be used during address screening.

Use Excluded Terms

• Use excluded terms (technical name EXCLUSION_TERMS) is a yes-no value that specifies whether the screening shall remove certain terms from names and addresses before comparing them. Excluded terms can be used to ignore common words that would otherwise produce lots of false positive. For example, the value Y makes the screening ignore the Ltd. in ABC Holdings Ltd. if it is entered as an excluded term. If a name or address consists of excluded terms only, for example Limited Corp., the excluded terms will be ignored to make the system produce any results at all.
• With the app Manage Excluded Terms you can create lists of terms to be excluded during address screening.

Use Initials of Name

Use initials of name (technical name INITIALS) is a yes-no value that specifies whether the screening shall consider one-letter abbreviations of names. For example, the value Y will match J. to James. Use this option to improve hit quality in countries/regions such as the United States, where initials are widely used.

List Type Group

List type group (technical name LIST_TYPE_GROUP) is the identifier of the list type group that the screening shall compare addresses against. While list ID directly identifies one of the lists delivered by the data provider, the list type group can be any combination of list segments from one or multiple such lists. The list type group therefore gives you more possibilities to recombine entities as needed.

Address Screening Type

• Address Screening Type (technical name SCREENING_TYPE) is a value field that specifies which parts of an address will be compared in order to produce matches. Value N (name only) makes the screening compare names only. That is, the address is completely ignored. Value C (name and country/region) requires an exact country/region match before names are compared. Value A (name and address) will compare the names and addresses, and O (address only) will only compare the address.
• For example, the value N will match a John Adams in London, UK to one in Washington, US. The value C will not make a match because the country/region is different. The value A will also not make a match because the address is different.

Address Screening Filter

The address screening filter (technical name ADDRESS_SCREENING_FILTER) has a value that identifies the name of the AMDP class used to filter out or reduce the number of address screening hits returned on the UI, by comparing additional attributes, such as gender or place of birth.

For example, the date of birth and gender can be additionally checked for natural persons to significantly reduce the number of false positives and improve quality of screening.

The following screening modes are supported by the address screening functionality:

The Online and Mass detection screening modes are technically same, as for other detection methods.

The Delta Screening mode is only used by address screening methods and only screens against list entities, imported with delta lists.

After this initial data load, you typically only receive any changes to those lists, i.e. the delta. To screen the entities in address screening lists that were changed since their initial import, perform the Delta Address Screening.

The delta screening can be either scheduled as a background job using the program FRA_ADDRESS_DELTA_SCREENING or initiated in ad hoc mode using the Detection Runs app.

Intelligent screening automatically closes new alerts if previous alerts with the same name and business address have been closed manually in the past.

New alerts in BIS that are created through address screening will be closed automatically if previous alerts with the same name and same business address (exact match) have been rejected by selecting NO in the hit toggle. If other types of detection methods were also used, the address screening hits with an exact match will be closed, but the alert items will remain open.

When an alert in BIS is created through address screening, the system will search for previous alert items with the same business address. If a previous alert item is found, the system will check the decision of each alert item. The highlighted name, highlighted address, country/region, and list classification are compared.

The closing reason that is required to close the alert items must be defined in Customizing activity Define Default Reasons for Closing Alerts.

Nothing. Intelligent screening automatically runs in your system. You just have to have had previous alerts with completion status False Alarm and address screening hits that were set to NO.

In additional to the automated screening of existing business partners, you can also perform a manual, ad hoc screening, of potential business partners by entering their name and address information into theAd Hoc Screening App.

Potential hits are displayed directly in the app and can be reviewed. In case of confirmed hits, the risk for the business partner can be assessed and respective risk remediation measures can be performed (for example the business partner onboarding can be stopped).

Please note, that no alerts are created for hits when using this app, also no audit trail records are created.

The audit trail logs all of the detection objects – persons, addresses, partners, and so on – that have been processed by address screening detection methods.

The purpose of the audit trail is to let you show your compliance with anti-corruption laws and regulations through your address screening activities and to help you analyze address screening hits. With the audit trail, you can show when a business partner was screened and with what result.

To enable the audit trail, set the Audit Trail flag to ON for each applicable address screening detection method.