Table 1. High-level feature variables... | Emerald Publishing

Table 1.

High-level feature variables

Types of variables	Rationale	Examples
Customer features	Unique customer elements could provide specific risk indicators associated with client’s demographics	Sector class, geographic location, number of years of client relationship, employment type, expected turnover, etc
Product use	Features related to various type of transactions identifies different risks	Cash intensive, high-risk countries, rapid movements, round amounts, channels, recency, risky merchant
Compliance features	Customers who are already part of any other investigation or watchlist or STR can be escalated to L2 directly	Previous STR, corr bank queries, watchlist, CIU or EDD concerns, payment declines
Relationships	Any unusual relationship based on counterparties or related parties	Number of counterparties, risky counterparty (STR), STR on related party
Existing AML rules	Indicators based on the past alerts and each detection scenario	DS flags, number of previous alerts

Types of variables	Rationale	Examples
Customer features	Unique customer elements could provide specific risk indicators associated with client’s demographics	Sector class, geographic location, number of years of client relationship, employment type, expected turnover, etc
Product use	Features related to various type of transactions identifies different risks	Cash intensive, high-risk countries, rapid movements, round amounts, channels, recency, risky merchant
Compliance features	Customers who are already part of any other investigation or watchlist or STR can be escalated to L2 directly	Previous STR, corr bank queries, watchlist, CIU or EDD concerns, payment declines
Relationships	Any unusual relationship based on counterparties or related parties	Number of counterparties, risky counterparty (STR), STR on related party
Existing AML rules	Indicators based on the past alerts and each detection scenario	DS flags, number of previous alerts

Source(s): Authors’ own creation/work