Georeferenced Transaction Data Clustering for POI Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Web-based mapping services face inaccuracies and incompleteness in point of interest (POI) data, leading to reliance on third-party vendors for updates, which limits control over the quality and quantity of data provided to client devices.
Innovation Solution
Georeferenced transaction data is harvested from client devices, analyzed to form clusters, and matched with supplemental data to update POI information, providing accurate and complete business information to users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If third-party vendors are used to update POI collections, then the quantity and quality of POI data can be increased, but control over the data provided to client devices is lost
Solution Approach 1:
The system enables self-service by allowing client devices to automatically contribute transaction data to the POI database. Users perform financial transactions with their mobile devices, and the system automatically harvests georeferenced transaction data from these devices to update POI information without requiring manual intervention from map service providers or reliance on third-party vendors.
2Quantity of substance
If commercial POI collections are purchased or obtained from royalty-free collections, then POI data availability is improved, but accuracy and completeness deteriorate
Solution Approach 1:
The system implements feedback by continuously harvesting transaction data from client devices and using this real-world usage information to update and improve POI database accuracy. The cluster analysis process analyzes georeferenced transaction data to identify patterns and update POI information, creating a feedback loop where actual transaction behavior refines the database quality over time.
Solution Approach 2:
The system replaces mechanical data collection methods (manual vendor updates, subscription-based POI collections) with automated electronic data harvesting from client devices. The system electronically captures transaction data directly from mobile devices, performs automated cluster analysis, and updates the POI database through electronic processes, eliminating the need for manual vendor interventions.
3Measurement precision
If cluster analysis is performed on georeferenced transaction data, then POI data accuracy is improved, but processing complexity increases
Solution Approach 1:
The system applies segmentation by dividing the georeferenced transaction data into clusters based on spatial and temporal patterns. The cluster analysis process segments transaction data into groups representing different POI locations and characteristics, making the complex data more manageable and enabling accurate POI identification through pattern recognition rather than processing all raw data uniformly.
Data Source
AI summary
Georeferenced transaction data is harvested (“crowd-sourced”) from client devices and sent to a network-based map service. The map service performs cluster analysis on location data points in the harvested data, resulting in one or more clusters representing local densities of transaction occurrences. Data vectors including supplemental data are obtained from one or more vendors. Location data points included in the data vectors are compared to center coordinates of the one or more clusters and the closest matching cluster/vector pair provides a mapping to POI data in a POI database. The mapped POI data is updated with the supplemental data. In some implementations, transaction timestamps in the harvested data are used to estimate the business hours of a business POI.