Transaction Classifier for User Location Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately determining user location from transaction records, leading to inefficient and inaccurate triggering of fraud prevention processes and targeted advertisements, as not all location information in transaction records is related to the user's location.
Innovation Solution
A method involving training a classifier using historical transaction data to classify whether a transaction description includes a user location, by assigning probability scores based on similarities with historical data, thereby distinguishing between user and business locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If location information from transaction records is used to trigger fraud prevention processes and targeted advertisements, then user experience can be improved through personalized services, but accuracy deteriorates because not all location information in transaction records is related to the user's location
Solution Approach 1:
The patent segments location information into two distinct categories: user location (derived from mobile device GPS coordinates) and business location (extracted from transaction records). This segmentation allows the system to use both types of location data appropriately - business location for targeted advertisements and user location for fraud prevention - thereby resolving the contradiction between service personalization and location accuracy.
Solution Approach 2:
The patent introduces an intermediary classification process that analyzes transaction descriptions to determine whether location information represents a user location or a business location. This intermediary step acts as a filter between the raw transaction data and the fraud prevention/advertisement systems, ensuring that only appropriate location data is used for each purpose, thus maintaining both service versatility and location precision.
2Reliability
If fraud prevention processes are triggered for all transactions with location information, then security can be enhanced, but resource efficiency deteriorates due to unnecessary processing of transactions where location info is not related to user location
Solution Approach 1:
The patent applies preliminary classification of transaction location information before triggering fraud prevention processes. By pre-analyzing whether the location in a transaction record represents the user's actual location or merely the business location, the system可以避免不必要的欺诈检测处理，从而在增强安全性的同时减少资源浪费。
Solution Approach 2:
The patent implements partial action by selectively applying fraud prevention processes only to transactions where location information indicates user presence. Rather than universally applying fraud detection to all transactions with location data, the system performs fraud prevention only when necessary (i.e., when the location corresponds to user location), thereby optimizing resource utilization while maintaining security effectiveness.
3Productivity
If targeted advertisements are provided based on transaction location information, then user engagement can be improved, but relevance deteriorates when the location information refers to business location rather than user location
Solution Approach 1:
The patent segments location data into user location and business location, enabling the advertisement system to selectively use business location information for targeted advertisements while using user location for fraud prevention. This segmentation ensures that advertisements are targeted based on relevant business context without compromising the accuracy of user location tracking, thereby maintaining both engagement and relevance.
Solution Approach 2:
The patent applies local quality by using different location information for different purposes: business location for targeted advertisements and user location for fraud prevention. This localized application of location data ensures that each function receives the most appropriate location context, improving advertisement relevance while maintaining fraud detection accuracy.
Data Source
AI summary
Certain aspects of the present disclosure provide techniques for improving a user experience based on electronic records of transactions. Embodiments include training a classifier using training data comprising a set of historical transaction descriptions and a set of corresponding historical classifications that indicate whether or not each historical transaction description of the set of historical transaction descriptions is associated with a user location. Embodiments further include receiving a transaction record describing a transaction associated with a user. Embodiments further include using the classifier to determine a classification for the transaction. The classification indicates whether or not the transaction description is associated with a location of the user. Embodiments further include providing, based at least in part on the classification, a communication to the user that relates to the location of the user.