Transaction Location Prediction via Consumption Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Transaction records often lack standardized formats and location information, making it inefficient and error-prone to determine locations for analysis, as conventional methods are slow, tedious, and inflexible, and often fail when dealing with varied formats or incomplete data.
Innovation Solution
A method that involves obtaining transaction records with user and merchant IDs, analyzing description strings to generate branch identification patterns, creating extended records by appending branch identifiers, constructing a consumption graph, and using this graph to estimate and refine location information through external APIs, thereby determining precise point locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional automated solutions rely on strict formatting requirements to determine location information, then processing speed is improved, but flexibility and accuracy deteriorate when analyzing transaction records of many different formats
Solution Approach 1:
The patent transforms the approach from using strict formatting rules to using machine learning models that can adapt to various data formats and patterns. The system changes the parameter of format recognition from rule-based to model-based, enabling it to handle diverse transaction record formats while maintaining processing efficiency.
Solution Approach 2:
The patent replaces conventional automated solutions that rely on mechanical rule-based formatting checks with machine learning-based pattern recognition. This substitution allows the system to automatically adapt to different formats without requiring explicit formatting rules, thereby maintaining both speed and flexibility.
2Measurement precision
If individual review and analysis of each transaction record is performed to determine location information, then accuracy is improved, but time consumption and labor intensity increase significantly
Solution Approach 1:
The patent implements a system where the machine learning model automatically extracts and determines location information from transaction records without requiring manual intervention. The model self-learns from training data and autonomously processes new records, achieving both high accuracy and efficiency.
Solution Approach 2:
The patent performs preliminary training of machine learning models using labeled transaction records with known location information. This preliminary action enables the model to learn patterns and features associated with location data, so that when processing actual transaction records, the system can quickly and accurately determine locations without manual review.
3Reliability
If only transaction records with complete location information are used for analysis, then data quality is improved, but the quantity of usable records decreases significantly
Solution Approach 1:
The patent introduces machine learning models as intermediaries between incomplete transaction records and location-based analysis requirements. The models infer missing location information by learning patterns from complete records, thereby enabling the use of incomplete records while maintaining data quality for analysis.
Solution Approach 2:
The patent changes the approach from requiring complete location information to using predictive modeling. By transforming the problem from data completion to pattern recognition, the system can reliably determine locations even when transaction records contain incomplete or missing location fields, thereby increasing the quantity of usable records.
Data Source
AI summary
Certain aspects of the present disclosure provide techniques for predicting a location based on transaction record data. An example technique includes obtaining a first set of transaction records and determining a merchant associated with each transaction record. The example further includes based on the merchant, determining and appending a branch identifier to each transaction record associated with the merchant to generate a first set of extended transaction records. The example further includes creating a consumption graph based on the first set of extended transaction records and determining an estimated location based on the consumption graph. The example further includes determining a precise point location based on the estimated location.