Feature-Based Autocorrection Using Spatial Model Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing auto-correction systems in computing devices often fail to detect non-spelling or non-grammar errors and may make erroneous corrections due to exact word matching, missing errors as long as the text includes a valid dictionary word, and incorrectly replacing intentionally typed non-dictionary words.
Innovation Solution
A method that uses a computing device to output a graphical keyboard, determine a character string, and assess a spelling probability based on spatial model probabilities to suggest corrections for potential spelling errors, allowing users to select accurate corrections efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If exact word matching is used for auto-correction, then dictionary words are correctly identified, but non-spelling errors and intentionally typed non-dictionary words cannot be detected
Solution Approach 1:
The system changes the detection parameter from exact word matching to probabilistic spelling probability assessment. Instead of checking if a word exists in the dictionary, the system calculates a spelling probability based on multiple features (spatial model, language model, character n-gram) to determine if the input is likely a misspelling, thereby detecting both spelling errors and intentional non-dictionary words
Solution Approach 2:
The auto-correction system becomes multi-functional by using a unified probabilistic framework that handles multiple error types (spelling errors, non-spelling errors, intentional non-dictionary words) through the same spelling probability assessment mechanism, rather than requiring separate detection systems for each error type
2Reliability
If traditional auto-correction systems are used, then spelling errors are detected, but the systems make erroneous corrections and fail to detect non-spelling or non-grammar errors
Solution Approach 1:
The system incorporates feedback from multiple sources (spatial model, language model, character n-gram) to continuously refine the spelling probability assessment. This multi-feedback mechanism improves correction reliability by cross-validating evidence before making corrections, reducing erroneous corrections while maintaining high detection accuracy
3Measurement precision
If multiple features are used to determine spelling probability, then detection accuracy is improved, but processing power consumption increases
Solution Approach 1:
The system applies partial action by selectively using multiple features only when needed for ambiguous cases. The spatial model is applied first as a quick filter, and additional features (language model, character n-gram) are applied only when the initial assessment is uncertain, thereby maintaining high accuracy while reducing overall processing power consumption
Data Source
AI summary
A computing device is described that outputs for display at a presence-sensitive screen, a graphical keyboard having keys. The computing device receives an indication of a selection of one or more of the keys. Based on the selection the computing device determines a character string from which the computing device determines one or more candidate words. Based at least in part on the candidate words and a plurality of features, the computing device determines a spelling probability that the character string represents an incorrect spelling of at least one candidate word. The plurality of features includes a spatial model probability associated with at least one of the candidate words. If the spelling probability satisfies a threshold, the computing device outputs for display the at least one candidate word.