Context Determination via Location Clustering and Probability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current context determination methods in mobile computing devices rely solely on sensory data, which can be inaccurate, especially in environments where environmental contexts are similar, such as home and office, leading to uncertain context identification.
Innovation Solution
The method predicts context by grouping locations with similar context histories into clusters and combining the probability of the predicted context with sensory data to determine the most likely context, using location-dependent context label distributions and clustering algorithms to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If context determination relies solely on sensory data, then the system simplicity is maintained, but the measurement precision deteriorates in similar environments
Solution Approach 1:
The system performs preliminary actions by collecting and storing historical context data during device usage, building a database of location-context associations before they are needed for prediction. This pre-processing enables more accurate context determination without adding complexity to the real-time decision-making process
Solution Approach 2:
Historical context data serves as an intermediary between raw sensory data and context determination. The system uses this intermediate layer of previously observed context patterns to enhance the accuracy of current context identification, particularly in ambiguous environments where sensory data alone is insufficient
2Measurement precision
If historical context data is used to predict context, then the context determination accuracy is improved, but the loss of time increases due to data processing
Solution Approach 1:
Location clusters and their associated context label distributions are pre-computed and stored during off-line training phases. When the device needs context prediction, the system simply retrieves pre-calculated probabilities for the current location cluster, avoiding time-consuming real-time analysis of historical data
Solution Approach 2:
The system serves itself by maintaining and updating its own historical context database automatically during normal operation. The context determination process uses this self-collected data to improve accuracy without requiring external intervention or complex real-time computations
Data Source
AI summary
Methods and apparatuses are provided for context determination. A method may include determining a location identifier. The location identifier may be indicative of a location of an apparatus. The method may further include determining a location cluster based at least in part on the location identifier. The method may additionally include determining a first probability for each of one or more contexts based at least in part on the determined location cluster. The method may also include determining a context of the apparatus based at least in part on the determined first probability. Corresponding apparatuses are also provided.