Journey Classification Model Personalization via Progressive Adaptation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users face difficulties in interpreting and acting on journey data recorded while moving in a vehicle, as existing technologies lack efficient methods for classifying the purpose of journeys.
Innovation Solution
A computer-implemented technique that automatically classifies journeys using a model-assisted classification procedure, generating features based on user data, and allowing user confirmation through a user interface, with progressive personalization of the machine-trainable model without retraining, enabling efficient use of resources and local processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a machine-trainable model is used to automatically classify journeys, then classification accuracy is improved, but the complexity of the system increases
Solution Approach 1:
The system dynamically adapts the classification model through progressive personalization, where the model evolves and refines its classification accuracy over time based on user feedback without requiring full retraining. This allows the system to maintain high accuracy while managing complexity through incremental adaptation rather than static complex model deployment.
Solution Approach 2:
The system changes parameters of the classification model progressively through user interactions. By adjusting model parameters based on confirmed classifications and user feedback, the system improves accuracy incrementally while avoiding the need to deploy a highly complex pre-trained model from the start.
2Use of energy by moving object
If progressive personalization is implemented without retraining the model, then processing resources are conserved, but the ability to adapt to new patterns may be limited
Solution Approach 1:
The system performs preliminary personalization by pre-processing user feedback and confirmed classifications to gradually adjust model parameters. This preliminary adaptation allows the system to learn user-specific patterns without requiring computationally expensive full retraining, balancing resource conservation with adaptation capability.
Solution Approach 2:
The classification model performs self-service by automatically adjusting its parameters based on user feedback and confirmed classifications. The system self-adapts to user patterns through progressive personalization without external retraining intervention, conserving processing resources while maintaining adaptability to new patterns.
3Loss of information
If journey data is processed locally on the user computing device, then data privacy is improved, but the computational power available is reduced
Solution Approach 1:
The system extracts only the essential classification functionality and model parameters to run locally on the user device. By separating the heavy retraining computations (performed elsewhere or incrementally) from the local inference and personalization operations, the system maintains data privacy while working within local computational constraints.
Solution Approach 2:
The system uses lightweight, simplified classification models that can be efficiently executed on resource-constrained mobile devices. These simplified models prioritize privacy by running locally while accepting reduced computational complexity, with the understanding that full model training occurs separately and only essential parameters are maintained locally.
4Productivity
If a sparse feature set is used, then processing efficiency is improved, but the detail and richness of journey characterization is reduced
Solution Approach 1:
The system extracts only the most discriminative and essential features from journey data for classification purposes. By selecting a sparse subset of key features that capture the most important patterns, the system achieves efficient processing while retaining sufficient information for accurate classification and personalization.
Data Source
AI summary
A technique is described herein for automatically logging journeys taken by a user, and then automatically classifying the purposes of the journeys. In one implementation, the technique obtains journey data from one or more movement-sensing devices as a user travels from a starting location to an ending location in a vehicle. The technique generates a set of features based on the journey data, and then uses a machine-trainable model (such as a neural network) to make its classification based on the features. The machine-trainable model accepts at least one feature that is based on statistical information regarding at least one aspect of prior journeys that the user has taken. Overall, the technique provides a resource-efficient solution that rapidly provides personalized results to individual respective users. In some implementations, the technique performs its personalization without sharing journey data with a remote server.