Predicting Escalation Events in Web Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Web-based searching often returns large volumes of medical information, some of which is erroneous, leading users to misinterpret common symptoms as serious illnesses, causing unnecessary anxiety and costly engagements with healthcare professionals.
Innovation Solution
A predictive model is constructed to estimate the likelihood of query escalation by extracting features from webpages and using a trained classifier to predict the probability of users escalating from common to more severe outcomes, providing an escalation likelihood score for subsequent search queries and webpage selections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If web-based searching returns large volumes of medical information, then information availability is improved, but users may misinterpret common symptoms as serious illnesses causing unnecessary anxiety
Solution Approach 1:
The patent introduces an intermediary system (search engine with predictive model) that mediates between the user's search query and the vast medical information available online. The system extracts features from webpages, runs them through a trained classifier to generate escalation likelihood scores, and uses these scores to predict when users might escalate from benign to severe outcome interpretations. This intermediary processing layer filters and contextualizes information before it reaches the user, preventing harmful misinterpretation while preserving information availability.
2Object-affected harmful factors
If a predictive model is constructed to predict query escalation, then user anxiety can be reduced, but system complexity increases
Solution Approach 1:
The patent segments the complex task of predicting query escalation into distinct components: feature extraction from webpages, classification using a trained model, escalation likelihood scoring, and prediction of subsequent user behavior. By dividing the system into these modular segments (feature extraction module, classification module, scoring module), the complexity is managed and made tractable while still achieving the goal of reducing user anxiety through accurate predictions.
Data Source
AI summary
Escalations in users' goals or concerns in web-based searching and browsing may be predicted. An escalation feature is extracted from a webpage and run through a classifier trained to estimate a likelihood that a subsequent search query will comprise an escalation when compared to a previous search query and/or that a subsequent webpage selection will comprise an escalation when compared to a previous webpage selection. It can thus be predicted whether a user visiting a current webpage is likely to escalate or navigate to another webpage based upon the current webpage, for example.