Prediction Engine for Mobile Devices Using User Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile devices require users to actively request information, such as routes, which can be inefficient given the vast amount of personalized data they can access, and there is a need for a system that can predict destinations and routes based on user-specific data without explicit user input.
Innovation Solution
A novel prediction engine that utilizes machine-learning to formulate predictions about current or future destinations and routes by analyzing user-specific data, including previous destinations, routes, calendared events, electronic tickets, and email addresses, and can relay this information to the user, either solely on the device or with external server data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If users actively request information (e.g., enter route requests), then the device can provide accurate information, but the operation becomes inefficient and time-consuming
Solution Approach 1:
The system performs preliminary actions by continuously analyzing user data (calendar events, past destinations, email addresses) to predict future destinations and routes before users explicitly request them. This allows the device to proactively prepare and present route information, eliminating the need for users to manually input requests and significantly reducing information retrieval time.
2Ease of operation
If the device leverages stored user-specific data to predict destinations and routes, then user convenience is enhanced, but the device complexity increases
Solution Approach 1:
The prediction engine is segmented into distinct functional modules: a data collection module that gathers user-specific information from multiple sources, a machine learning module that analyzes patterns in the collected data, and a route generation module that creates predicted destinations and routes. This modular segmentation manages complexity by allowing each component to be developed and optimized independently while working together to provide automated predictions.
3Measurement precision
If the prediction engine runs frequently to provide accurate predictions, then prediction accuracy improves, but energy consumption increases
Solution Approach 1:
The prediction engine implements periodic action by running predictions at scheduled intervals rather than continuously. The system analyzes user data and generates route predictions at specific times (e.g., morning commutes, evening returns) based on learned patterns, rather than constantly monitoring and processing data. This periodic operation maintains prediction accuracy for frequently traveled routes while significantly reducing overall energy consumption compared to continuous operation.
Data Source
AI summary
Some embodiments of the invention provide a novel prediction engine that (1) can formulate predictions about current or future destinations and/or routes to such destinations for a user, and (2) can relay information to the user about these predictions. In some embodiments, this engine includes a machine-learning engine that facilitates the formulation of predicted future destinations and/or future routes to destinations based on stored, user-specific data. The user-specific data is different in different embodiments. In some embodiments, the stored, user-specific data includes data about any combination of the following: (1) previous destinations traveled to by the user, (2) previous routes taken by the user, (3) locations of calendared events in the user’s calendar, (4) locations of events for which the user has electronic tickets, and (5) addresses parsed from recent e-mails and/or messages sent to the user. In some embodiments, the prediction engine only relies on user-specific data stored on the device on which this engine executes. Alternatively, in other embodiments, it relies only on user-specific data stored outside of the device by external devices/servers. In still other embodiments, the prediction engine relies on user-specific data stored both by the device and by other devices/servers.