Social Media Adverse-Effect Detection Using Healthcare Taxonomy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently and automatically identifying potential adverse effects from healthcare products in vast amounts of social media and mobile application data, making manual monitoring impractical due to the sheer volume of content produced.
Innovation Solution
A system and method utilizing a healthcare taxonomy and linguistic rules to analyze online postings, score them based on relevance and veracity, and flag potential adverse effects, with human expert review and notification mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual monitoring of social media and mobile application data is performed to identify adverse effects, then detection accuracy can be maintained through human expert review, but the sheer volume of content produced makes manual monitoring impractical and time-consuming
Solution Approach 1:
The patent introduces an automated text mining system with natural language processing capabilities as an intermediary between social media data and human experts. This intermediary automatically processes vast amounts of UGC, applies linguistic rules and healthcare taxonomies to identify potential adverse effects, and pre-screens reports before human review, thereby maintaining detection accuracy while dramatically reducing the time investment required from human monitors
Solution Approach 2:
The patent replaces the mechanical system of manual human monitoring with an automated computational system that uses text mining algorithms, natural language processing, and scoring mechanisms. This substitution enables the system to process large volumes of social media content automatically, identifying and flagging potential adverse effects without requiring continuous human time investment while maintaining or improving detection accuracy through systematic analysis
2Productivity
If automated text mining and natural language processing are used to analyze online postings, then the volume of data that can be processed increases significantly, but the complexity of the system increases with multiple components including taxonomy databases and linguistic rules
Solution Approach 1:
The patent segments the complex task of adverse effect detection into distinct modular components: (1) data collection from multiple sources, (2) text mining with natural language processing, (3) scoring based on multiple criteria (proximity of keywords, discussion context, specificity, relevancy degree), (4) threshold comparison and flagging, and (5) human expert review. This segmentation allows each component to be independently optimized and managed, enabling high data processing capacity while making the overall system complexity tractable through modular architecture
Solution Approach 2:
The patent creates a multi-functional system that simultaneously performs data collection from diverse sources (social networking sites, discussion forums, app stores), text mining, natural language processing, error correction, scoring, and flagging operations. The healthcare taxonomy and linguistic rules serve universal purposes across all analysis functions, allowing the system to handle various types of healthcare product discussions with a single unified framework, thereby achieving high productivity without proportionally increasing complexity
3Measurement precision
If scoring mechanisms are implemented to evaluate the likelihood of adverse effects based on multiple criteria, then the precision of adverse effect identification improves, but additional computational resources and processing time are required
Solution Approach 1:
The patent implements a tiered scoring approach where not all postings receive full multi-criteria analysis. Instead, the system applies quick preliminary filters to identify obviously relevant or irrelevant content, then applies the full scoring mechanism (proximity of keywords, discussion context, specificity, relevancy degree) only to borderline cases that require more sophisticated evaluation. This partial application of the complex scoring mechanism maintains high identification precision for critical cases while reducing overall computational resource consumption
Data Source
AI summary
Some implementations provide a computer-implemented method for identifying, from on-line postings, reports of potential adverse effects resulting from consuming a healthcare product, the method including: receiving a log of on-line postings regarding consuming the healthcare product; receiving a database comprising a healthcare taxonomy and a set of linguistic rules; analyzing, based on the healthcare taxonomy, the log of on-line postings to identify a report of at least one adverse effect resulting from consuming the healthcare product; generating a score for the identified report according to the healthcare taxonomy and the set of linguistic rules; comparing the generated score with a threshold; and in response to determining that the generated score is above the threshold, flagging the identified report as a report of a potential adverse effect.