AI Device Sampling with Static and Dynamic Weights
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing number of electronic devices in the marketplace makes it difficult and expensive to understand user usage behavior from usage data, as existing sampling methods often only select active devices, failing to represent the overall activity of all devices, including both active and inactive ones.
Innovation Solution
An artificial intelligence-based multi-goal-aware device sampling method that determines static and dynamic weights for electronic devices from multiple data sources, identifying a representative subset of devices that includes both active and inactive ones to accurately reflect overall device activity across various demographic categories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sampling methods are used that only select active devices, then the sampling process is simple and fast, but the representativeness of the device population is poor and cannot reflect overall device activity including inactive devices
Solution Approach 1:
The patent applies dynamics by transitioning from static sampling (only active devices) to dynamic sampling that continuously adjusts device selection based on activity states. The system dynamically reweights devices as they transition between active and inactive states, ensuring the sample pool adapts to current population characteristics while maintaining computational feasibility through incremental updates rather than complete recalculation
Solution Approach 2:
The patent changes the parameter of device selection criteria from binary (active/inactive) to continuous weighting factors. By introducing activity-based weights that can take any value between 0 and 1, the system transforms the sampling approach to include inactive devices with reduced weight, thereby improving representativeness while managing complexity through mathematical weighting rather than complex selection logic
2Loss of information
If comprehensive device sampling including inactive devices is implemented, then the understanding of overall device activity improves, but the computational cost and processing complexity increase
Solution Approach 1:
The patent extracts the essential characteristic (device activity level) from the complex data and uses it as a weighting factor. By separating the activity assessment from the sampling selection process, the system can efficiently compute weights based on simple activity metrics while using these weights to guide more complex sampling decisions, thereby reducing overall processing complexity
Solution Approach 2:
The patent introduces activity weights as an intermediary variable between raw device data and sampling selection. This intermediary transforms complex device states into simple numerical weights that can be efficiently processed and used to probabilistically select devices, including inactive ones, without requiring complex processing of the underlying device data
3Productivity
If static weights alone are used for device sampling, then the sampling process is computationally efficient, but the ability to adapt to changing device activity patterns is limited
Solution Approach 1:
The patent implements dynamics by combining static baseline weights with dynamic activity-based weights. The static component ensures computational efficiency and provides a stable foundation, while the dynamic component continuously adapts to changing device activity patterns. This hybrid approach allows the system to maintain high sampling efficiency while being responsive to temporal changes in device usage
Solution Approach 2:
The patent applies preliminary action by pre-computing static weights based on historical device characteristics and usage patterns. These pre-computed weights serve as a foundation that reduces the computational burden during real-time sampling, allowing the system to quickly adapt to changes by adjusting only the dynamic activity component rather than recalculating all weights from scratch
Data Source
AI summary
An electronic device includes at least one processor configured to obtain user data associated with a plurality of devices from multiple data sources. The at least one processor is also configured to determine a static weight for each of the plurality of devices based on at least one source of the multiple data sources. The at least one processor is further configured to identify a portion of the plurality of devices that represents the plurality of devices based on the static weight and a dynamic weight. In addition, the at least one processor is configured to determine the dynamic weight for each of the portion of the plurality of devices while the portion of the plurality of devices is identified, where the dynamic weight is based on one or more sources of the multiple data sources.