Chinese Text Error Correction via Glyphic Phonetic Coding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Chinese input methods, particularly those based on phonetic readings like Pinyin and root shapes, are prone to errors due to typing inaccuracies, homophone confusion, and systemic biases, especially in automated speech recognition and handwriting recognition systems, which struggle to accurately process the complex glyphic and phonetic features of Chinese characters.
Innovation Solution
A text error identification and correction system that utilizes a glyphic and phonetic coding (GPC) system to encode Chinese characters, allowing for flexible similarity calculations and candidate word ranking, incorporating deep neural networks and domain-specific dictionaries to correct errors by converting text inputs into fixed-length sequences for phonetic and glyphic components.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If phonetic-based input methods (Pinyin) are used for Chinese text input, then typing speed and ease of input are improved, but typing errors and homophone confusion increase
Solution Approach 1:
The patent introduces an intermediary error correction system that mediates between the phonetic input method and the final text output. The system captures phonetic input, encodes it with glyphic information, identifies potential errors through similarity comparison, and corrects them before final output, thus resolving the contradiction between ease of phonetic input and typing accuracy
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors input text for potential errors by comparing phonetic and glyphic features, provides error identification to the user, and allows correction based on candidate suggestions. This feedback loop enables users to maintain the ease of phonetic input while improving overall typing accuracy through systematic error detection and correction
2Productivity
If automated speech recognition systems process Chinese characters, then processing speed is improved, but accuracy deteriorates due to complexity of glyphic and phonetic features
Solution Approach 1:
The patent segments the Chinese character processing into distinct phonetic and glyphic components, encoding each separately. This segmentation allows the automated system to process phonetic information quickly while independently verifying glyphic accuracy, thus maintaining both processing speed and recognition accuracy despite the complexity of Chinese characters
Solution Approach 2:
The patent changes the parameter representation of Chinese characters by encoding them as fixed-length sequences that combine phonetic and glyphic features. This parameter transformation enables automated speech recognition systems to efficiently process Chinese text while maintaining accuracy through the structured encoding scheme that preserves both phonetic and visual character information
3Productivity
If root shape input methods are used for Chinese text input, then typing speed is improved, but difficulty in learning and usage increases
Solution Approach 1:
The patent introduces an intermediary error correction system that mediates between root shape input and final text output. The system captures root shape input, encodes it with phonetic information, identifies potential errors through similarity comparison, and corrects them automatically or with minimal user intervention, thus resolving the contradiction between typing speed and ease of usage
Data Source
AI summary
A method, computer program, and computer system is provided for text error identification and correction. A text input having a phonetic component and a glyphic component is received. Information corresponding to the phonetic component and the glyphic component is coded as a fixed-length sequence. One or more candidate replacement words corresponding to the fixed-length sequence are identified. At least a portion of the text input is replaced with a candidate replacement word from among the one or more candidate replacement words.