Electronic Device Activity Data Clustering and Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Electronic devices face challenges in verifying the reliability of user activity data, which can lead to incorrect or erroneous services being provided without a separate verification procedure, especially as large volumes of user activity data are collected and analyzed.
Innovation Solution
An electronic device comprising sensors, a processor, and memory that transmits sensing data to a server for clustering and similarity analysis, using center similarity, variance, and intersection scores to identify the reliability of new data and execute functions accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If large volumes of user activity data are collected to improve service accuracy, then the quantity and comprehensiveness of data increase, but the difficulty of verifying data accuracy and reliability increases
Solution Approach 1:
The system performs preliminary clustering of user activity data into multiple clusters representing different activity patterns before verification. This preliminary organization allows for more efficient accuracy checking by comparing new data against pre-established activity patterns rather than verifying each data point individually, thus maintaining reliability while handling large data volumes
Solution Approach 2:
The patent introduces cluster centers as intermediary representations of activity patterns. Instead of directly verifying raw sensing data against ground truth, the system uses cluster centers as mediators to represent typical activity patterns. New data is verified by comparing its similarity to these intermediary cluster representations, simplifying the verification process for large datasets
2Measurement precision
If multiple clusters are used to represent different user activities, then the ability to accurately categorize and verify data improves, but the complexity of data processing and analysis increases
Solution Approach 1:
The patent segments user activity data into multiple distinct clusters, each representing a specific type of activity pattern. This segmentation allows the system to handle complex diverse activities by breaking them down into manageable categories, improving measurement precision while keeping individual cluster processing relatively simple
Solution Approach 2:
The system uses similarity scores as a key parameter to simplify cluster comparison. Instead of complex multi-dimensional comparisons, the patent transforms the complexity into a single similarity metric that quantifies how well new data matches each cluster, significantly reducing processing complexity while maintaining accuracy
3Reliability
If similarity analysis based on multiple scores (center similarity, variance, intersection) is performed, then the reliability of data verification improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent calculates multiple similarity scores (center similarity, variance-based score, intersection-based score) but applies them selectively. The system computes these scores only when needed for verification decisions, and uses thresholds to determine when full multi-score analysis is necessary versus when simpler checks suffice, balancing reliability with processing time
Solution Approach 2:
The system uses the results of similarity score calculations as feedback to adjust processing. When similarity scores indicate clear cluster membership, the system can make quick verification decisions. When scores are ambiguous or close to thresholds, the system intensifies analysis by computing additional scores, thus optimizing processing time while maintaining verification reliability
Data Source
AI summary
According to certain embodiments, an electronic device comprises a communication module; a plurality of sensors and configured to obtain sensing data; at least one processor operatively connected to the plurality of sensors and the communication module; and a memory operatively connected to the at least one processor, wherein the memory stores instructions that, when executed, cause the at least one processor to perform a plurality of operations comprising: transmitting the sensing data to a server through the communication module; receiving, from the server, information on a similarity between the sensing data and a first cluster among a plurality of clusters clustering data related to user activities, through the communication module, wherein the similarity is identified based on a center similarity score between the sensing data and the first cluster, a score that is a function of a variance of the first cluster, a score that is a function a distance between the first cluster and other clusters, and an intersection score between the first cluster and a second cluster adjacent to the first cluster; and executing a function corresponding to the sensing data based on the similarity.