Virtual Keyboard Drag Gesture Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for text input on small devices like smartphones and smart watches are inefficient due to the difficulty in correcting errors, as they require multiple actions such as moving the cursor and deleting characters, which can be challenging on small screens, and auto-correction algorithms often fail to correct mistakes when both intended and typed strings are valid words or not in the dictionary.
Innovation Solution
A system that allows users to correct text input errors by dragging their finger from an incorrect character to the correct character on a virtual keyboard, replacing the incorrect character in the string, and displaying the corrected string, with options for inserting, deleting, or swapping characters, and utilizing error correction modes to facilitate easy correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional cursor-based correction methods are used on small devices, then correction functionality is available, but the number of actions required increases and ease of operation deteriorates
Solution Approach 1:
The patent combines multiple correction actions (cursor movement, character deletion, and character insertion) into a single drag gesture operation. When a user drags from one character to another on the virtual keyboard, the system simultaneously identifies the erroneous character, removes it, and inserts the correct character in one continuous motion, thereby reducing the number of discrete actions required.
Solution Approach 2:
The patent introduces a drag gesture as an intermediary input method that mediates between the user's intent to correct an error and the system's correction actions. The drag gesture serves as a unified interface that communicates both the location of the error and the desired correction to the system without requiring separate cursor movements and key presses.
2Reliability
If auto-correction algorithms are used, then some errors are corrected automatically, but correction accuracy deteriorates when both strings are valid words or not in dictionary
Solution Approach 1:
The patent enables users to manually initiate and control the correction process through drag gestures, allowing them to override or supplement auto-correction algorithms. Users can directly indicate corrections by dragging from erroneous to correct characters, giving them full control in ambiguous cases where auto-correction may fail or produce incorrect results.
Solution Approach 2:
The system provides visual feedback by displaying the erroneous string and allowing users to see the effect of their drag gestures in real-time. This feedback mechanism enables users to verify corrections and make adjustments if needed, improving overall correction accuracy especially in cases where auto-correction is uncertain.
3Ease of operation
If virtual keyboards are used on small devices, then text input is enabled, but the size of keys decreases making precise input more difficult
Solution Approach 1:
The patent adds a temporal and spatial dimension to key selection by allowing users to drag across multiple keys in a single continuous gesture. Instead of requiring precise point-and-tap input on small keys, users can sweep their finger across the keyboard surface, making the selection process more tolerant of imprecision and easier to perform with larger fingers or while moving.
Data Source
AI summary
An input signal corresponding to an action other than a drag on any virtual keyboard causes an erroneous string having at least two characters with an incorrect character other than the last character to be displayed. A second input signal corresponding to a drag on a virtual keyboard triggers entry into an error correction mode. A first incorrect character is located, a corrected input is determined according to an angle and a slide direction of the first drag, and the layout and geometry of the virtual keyboard; the first incorrect character is replaced with the corrected input to provide and display a first corrected string. The replacement of the first incorrect character and the display of the first corrected string occur without input from any source external to the device other than the first and the second input signals.