Routine Estimation Using Time-Based Geo-Location Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current social-networking systems face challenges in accurately predicting user location routines and detecting deviations from these routines without manual input, leading to inefficiencies in location data processing and notification systems.
Innovation Solution
The system groups ambient-location readings by time of day and day of the week, calculates probabilities of user presence at specific locations, and adjusts polling frequency based on inferred routines, sending notifications when deviations are detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system continuously monitors location data to accurately predict user routines, then the accuracy of location routine inference is improved, but the amount of location data processing increases
Solution Approach 1:
The system performs preliminary actions by grouping location readings into time-based clusters before analysis. It pre-processes the data by organizing readings into hourly and daily clusters, which simplifies the subsequent routine detection process and reduces the computational burden on the full dataset.
Solution Approach 2:
The patent segments location data into distinct time-based clusters (hourly clusters and daily clusters). This segmentation allows the system to analyze patterns at different temporal granularities separately, improving measurement precision while reducing the overall quantity of data that needs processing by focusing on significant temporal patterns.
2Loss of time
If the system processes location data frequently to detect routine deviations, then the timeliness of notifications is improved, but the processing overhead increases
Solution Approach 1:
The system implements periodic action by continuously updating routine predictions based on newly arrived location readings. It periodically re-evaluates clusters and compares them against established routines to detect deviations, enabling timely notifications while managing processing overhead through structured periodic updates rather than continuous intensive processing.
Solution Approach 2:
The patent employs feedback mechanisms where the system monitors actual user location against predicted routines and adjusts its models accordingly. When deviations are detected, the system provides feedback through notifications to connected users, creating a closed-loop system that improves timeliness while controlling processing complexity through targeted updates.
Data Source
AI summary
In one embodiment, a method includes determining a geo-location centroid of each of one or more geo-location clusters. The geo-location centroid corresponds to one or more geo-location data points within its geo-location cluster. The geo-location data points represent one or more location readings from a mobile computing device associated with a user. The geo-location centroids are based at least in part on location readings obtained during a particular time of day of a particular day of a week. The method also includes grouping one or more geo-location centroids into one or more groups; and determining a time-based routine based at least in part on a number of geo-location centroids within each group.