Data Transaction Profile Compression via Bit String Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fraud detection systems face challenges in efficiently handling richer profiles due to fixed profile sizes, which limits their ability to counter sophisticated fraudulent transactions without significant modifications or upgrades.
Innovation Solution
The compression of data transaction history profiles using bit strings that characterize real numbers, allowing for reduced precision storage and subsequent decompression, enabling the use of richer profiles without altering existing systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If fixed profile size is used in fraud detection systems, then system simplicity and compatibility are maintained, but the ability to handle richer profiles and counter sophisticated fraud is limited
Solution Approach 1:
The patent changes the parameter representation from fixed-size floating-point numbers to variable-bit integer representations. By transforming profile variables into compressed integer formats with adjustable bit widths, the system can accommodate richer profiles while maintaining compatibility with existing fixed-size infrastructure through configurable compression parameters.
Solution Approach 2:
The patent introduces dynamic profile sizing capability where the effective profile size can be adjusted based on transaction complexity and fraud risk. The system can dynamically allocate more bits for high-risk transactions requiring richer profiles, while using minimal bits for low-risk transactions, making the profile size flexible rather than fixed.
2Reliability
If richer profiles are implemented to counter sophisticated fraud, then fraud detection capability is improved, but system upgrades and modifications are required
Solution Approach 1:
The patent creates a compressed representation layer that copies the essential information from complex floating-point profiles into simplified integer formats. This compression layer acts as an intermediary that preserves fraud detection capability while reducing storage and processing requirements, allowing richer profiles to be implemented without proportional increases in system complexity.
Solution Approach 2:
The patent segments the profile data into multiple compressed integer fields with different bit widths assigned to different profile variables. This segmentation allows selective allocation of precision and storage resources, enabling richer profiles for critical fraud indicators while maintaining simpler representations for less critical variables, thereby improving fraud detection without requiring uniform system upgrades.
3Measurement precision
If profile variables are stored with full precision, then measurement precision is maintained, but storage space and processing resources are consumed unnecessarily
Solution Approach 1:
The patent changes the storage parameter from fixed-precision floating-point to variable-precision integer representations. By transforming profile variables into compressed integer formats with adjustable bit widths, the system can maintain sufficient measurement precision for fraud detection while significantly reducing storage space consumption through configurable compression parameters.
Solution Approach 2:
The patent applies partial compression where not all profile variables require the same level of precision. The system can allocate more bits to variables critical for fraud detection (excessive action for those variables) while using minimal bits for less critical variables, achieving sufficient overall precision without the cost of full precision everywhere.
Data Source
AI summary
Techniques for compressing data transaction history profiles are disclosed. Such profiles can include a plurality of profile variables with each profile variable comprising a real number that provides a factor for determining whether a proposed data transaction is indicative of fraud. A bit string is generated for each profile variable in the profiles that characterizes a first value plus a second value. The first value is equal to a mantissa for the real number corresponding to the profile variable. The second value is equal to a number of orders of magnitude above a minimum required expressed as multiples of the number of orders magnitude required to represent the plurality of real numbers in each the plurality of transaction history profiles divided by a range of bits. The generated bit string is stored as compressed profile variable within the data transaction history profiles. Related systems, apparatus, methods, and/or articles are also described.