Text Prediction Engine Combining Multi-Attempt Evidence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Text prediction engines often fail to accurately predict user intent, leading to the display of unwanted candidates or incorrect auto-corrections, which can result in users having to manually edit or re-enter text.
Innovation Solution
A computing device with a text prediction engine that combines evidence from a first and second user attempt at inputting text to generate improved candidate lists, promoting matching candidates and filtering non-matching ones to provide more accurate predictions and corrections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a text prediction engine generates candidates based on a single user attempt, then the system operates quickly with simple processing, but the prediction accuracy is insufficient leading to unwanted candidates or incorrect auto-corrections
Solution Approach 1:
The system performs preliminary text prediction based on the first user attempt, generates candidate lists, and stores them for later use. When a second attempt is detected, the pre-generated candidates from the first attempt are retrieved and combined with new candidates, avoiding the need to generate everything from scratch and improving accuracy without proportional complexity increase
Solution Approach 2:
The system uses the outcome of the first user attempt (including which candidates were presented and whether correction was needed) as feedback to improve subsequent predictions. The evidence from the first attempt is stored and used to inform the prediction model during the second attempt, creating a feedback loop that progressively improves accuracy
2Ease of operation
If the text prediction engine presents unwanted candidates or makes incorrect auto-corrections, then the user must manually edit or re-enter text, which increases user effort and time consumption
Solution Approach 1:
The system merges candidate lists from multiple user attempts by combining evidence from the first attempt with evidence from the second attempt. This creates a more comprehensive candidate set that leverages information from both attempts, improving prediction reliability and reducing the need for manual correction while maintaining ease of operation
Data Source
AI summary
The description relates to predicting text based on multiple user attempts at inputting text. One example can include a computing device comprising a user interface. In this example, the user interface is configured to receive evidence from a first user attempt at inputting text and configured to receive evidence from a second user attempt at inputting the text. The computing device also includes a text prediction engine configured to combine the evidence from the first user attempt at inputting the text and the evidence from the second user attempt at inputting the text to predict the text.