Digital Memory Model for Scalable Human Behavior Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for modeling a human are activity-specific and require complex calibration against a large population baseline, making them non-scalable and not generally applicable across various physical and digital applications.
Innovation Solution
A system that uses sensors to gather personal data, processes it to generate digital memories through augmentation and analysis, and maintains associations between these memories, enabling a generic and scalable mechanism for modeling a human, which can be used in various applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If activity-specific modeling techniques are used, then the model can be calibrated for a specific task, but the model cannot be generally applied to other areas of activity
Solution Approach 1:
The patent creates a universal digital memory model that can serve multiple functions across different activities. The model stores personal data, event information, and associations in a structured format that can be applied to various domains including health, education, entertainment, and productivity, eliminating the need for separate activity-specific models while maintaining accuracy through personalized data processing
2Measurement precision
If complex calibration against large population baseline is performed, then the model can be accurately calibrated, but the process becomes time-consuming and complex
Solution Approach 1:
The patent extracts and processes only the relevant personal data and event information needed for the specific individual's model, eliminating the need for complex calibration against large population baselines. The system directly processes personalized data from sensors and inputs to create accurate digital memories without time-consuming population comparisons
Solution Approach 2:
The system performs preliminary data processing and structuring of personal information before model creation. By pre-processing sensor data, event information, and personal characteristics into a standardized format, the system eliminates the need for subsequent complex calibration processes while maintaining high accuracy
3Reliability
If regular validation is performed to reflect evolution of baseline and individual, then the model remains accurate, but the process requires continuous maintenance and validation
Solution Approach 1:
The digital memory model performs self-validation by continuously processing new personal data and event information to update its own structure. The system automatically adapts to changes in the individual's baseline and preferences without requiring external validation processes, maintaining accuracy through continuous self-adjustment based on new data inputs
Data Source
AI summary
A technique is provided for creating digital memories for a particular person. A data store stores personal data derived from signals gathered from a plurality of sensors that monitor the particular person. Memories creation processing circuitry, responsive to detection of a given event associated with the particular person, performs an augmentation process to generate an augmented given event identifying multiple items of data associated with the given event, including personal data associated with the given event obtained from the data store. The memories creation processing circuitry analyses the multiple items of data identified by the augmented given event in order to generate a given digital memory for the given event. A memories data store stores digital memories generated by the memories creation processing circuitry for the particular person, and memories analysis circuitry determines and maintains associations between the digital memories in the memories data store. Digital twin creation circuitry may then develop cognitive skills via analysis of the digital memories and their associations, for use in assisting the particular person.