Free-Text Event Detection for Digital Therapeutic Applications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to efficiently detect and manage adverse events in digital therapeutic applications due to the challenges of user feedback processing, particularly in free text formats from various channels, leading to inefficiencies in event monitoring and reporting, which affects user experience and compliance with regulatory requirements.
Innovation Solution
A machine learning-based system that processes free text from multiple channels using NLP and classifier models to detect events, generate analytics reports, and automate responses, reducing manual intervention and enhancing compliance with regulatory reporting.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review of user feedback submissions is used to detect events, then accuracy in determining whether users experienced events can be maintained, but the process becomes time-consuming and cannot scale to large volumes of users across multiple sources
Solution Approach 1:
The patent introduces machine learning models as an intermediary between user feedback submissions and administrator review. The ML models automatically analyze free text from multiple channels (emails, chat logs, call transcripts), extract event indicators, and prioritize submissions for human review. This intermediary layer maintains detection accuracy by using trained models while dramatically improving processing throughput and scalability.
Solution Approach 2:
The patent replaces the manual mechanical review process with an automated machine learning-based system. The ML architecture processes free text submissions, identifies adverse events, and generates structured outputs without requiring manual reading of each submission. This substitution maintains precision through trained model accuracy while eliminating the time and resource constraints of manual processing.
2Ease of operation
If digital therapeutics allow users to access treatment outside clinical environments, then user convenience and accessibility are improved, but the ability to monitor and detect adverse events becomes significantly more difficult
Solution Approach 1:
The patent implements a feedback mechanism where users continuously submit free text feedback through multiple channels (in-app messaging, emails, chat logs) during remote treatment. The machine learning system analyzes this ongoing feedback stream to detect adverse events, enabling monitoring without requiring clinical presence. The feedback loop maintains accessibility while solving the monitoring challenge through automated analysis of user-generated content.
Solution Approach 2:
The patent introduces machine learning analysis as an intermediary between remote users and clinicians. The ML system acts as a virtual monitor that continuously processes user feedback from various channels, detects adverse events, and alerts appropriate personnel. This intermediary enables the system to maintain the benefits of remote access while restoring adverse event detection capabilities that would otherwise require direct clinical observation.
3Quantity of substance
If multiple feedback channels are used to collect user experiences, then comprehensive data collection is improved, but the complexity of processing and analyzing free text from diverse sources increases
Solution Approach 1:
The patent implements a universal machine learning processing architecture that handles multiple feedback channel types (emails, chat logs, call center transcripts, in-app messages) through a single unified system. The ML models are trained to recognize event indicators across different formats and channels, eliminating the need for separate processing pipelines for each source. This multi-functional approach increases data collection comprehensiveness while managing processing complexity through standardization.
Data Source
AI summary
Aspects of the present disclosure are directed to systems, methods, and computer readable media for executing actions for events associated with use of applications. A computing system can identify data associated with an application to be evaluated for at least one of a plurality of events associated with a use of the application. The computing system can determine, based on applying the free text to a machine learning (ML) architecture, a value indicating a likelihood of occurrence of an event associated with the use of the application. The computing system can provide to a generative ML model, a model input based on the free data and the value. The computing system can execute an action corresponding to characterizing the event.