Phonetic Annotation for Polyphonic Chinese Text

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Learners of Chinese face obstacles in pronouncing polyphonic word segments while reading Chinese character texts aloud, as existing methods require significant time to look up pronunciations and often result in inaccuracies due to large vocabulary sizes, leading to inefficiencies in Chinese learning.

Innovation Solution

A method and apparatus for phonetically annotating Chinese character texts that identify polyphonic word segments, predict probabilities of their pronunciations, and select the most appropriate pronunciation based on a predetermined threshold, using a Conditional Random Fields (CRF) model and pre-established dictionaries, to improve accuracy and efficiency in text reading.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If learners look up pronunciations in a dictionary, then they can obtain pronunciation information, but it requires significant time and is difficult to grasp the most accurate pronunciation

Engineering Contradiction:

Improvepronunciation accuracyVSAvoidtime required for lookup

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system pre-processes and stores pronunciation probability information for polyphonic word segments in a database before actual use. When a word segment is encountered during text reading, the pre-calculated pronunciation probabilities are immediately retrieved and used, eliminating the need for real-time dictionary lookup and enabling instant pronunciation determination with high accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The manual dictionary lookup process is replaced by an automated electronic system that uses probabilistic models and database queries to determine pronunciations. The system automatically identifies polyphonic word segments, retrieves pre-stored pronunciation probabilities, and selects the most appropriate pronunciation without requiring manual intervention or time-consuming search processes

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Adaptability or versatility

If a large vocabulary dictionary is used, then more word segments can be covered, but it becomes difficult for users to grasp the most accurate pronunciation

Engineering Contradiction:

Improvevocabulary coverageVSAvoidpronunciation accuracy

Core Design Contradiction:

Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

Instead of treating all pronunciations equally across the large vocabulary, the system assigns different probability weights to different pronunciations of polyphonic word segments based on contextual analysis. This allows the system to handle diverse vocabulary while providing locally optimized, context-appropriate pronunciation selections with high accuracy for each specific word segment

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The system transforms the static dictionary lookup approach into a dynamic probabilistic selection process. By introducing probability parameters that reflect contextual appropriateness, the system can efficiently handle large vocabulary coverage while maintaining high pronunciation accuracy through quantitative comparison of pronunciation probabilities

Inventive Principle:

Principle #35Parameter changes

3Reliability

If manual dictionary lookup is used for each polyphonic word segment, then pronunciation can be obtained, but the efficiency and effects in text reading are lowered

Engineering Contradiction:

Improvepronunciation correctnessVSAvoidtext reading efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system enables continuous, uninterrupted text reading by automatically determining pronunciations on-the-fly without requiring pauses for dictionary lookup. The probabilistic pronunciation selection operates continuously as text is processed, maintaining both high pronunciation correctness and reading flow efficiency

Inventive Principle:

Principle #20Continuity of useful action

Solution Approach 2:

The system performs self-service by automatically identifying polyphonic word segments and selecting appropriate pronunciations without requiring user intervention. This automated process maintains pronunciation correctness while significantly improving text reading efficiency compared to manual lookup methods

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS10114809B2Method and apparatus for phonetically annotating text

Publication Date: 2018.10.30 TENCENT TECHNOLOGY (SHENZHEN) CO LTD

AI summary

Method for phonetically annotating text is performed at a computing device. The method includes: identifying a first polyphonic word segment in a text input, the first polyphonic word segment having at least a first pronunciation and a second pronunciation; determining at least a first probability for the first pronunciation and a second probability for the second pronunciation; determining a predetermined threshold difference based on: a comparison of the first and second probabilities with a preset threshold probability value, respectively, and a magnitude of a difference between the first and second probabilities; comparing the difference between the first probability and the second probability with the predetermined threshold difference; and selecting the first pronunciation as a current pronunciation for the first polyphonic word segment in accordance with a determination that the difference between the first probability and the second probability exceeds the predetermined threshold difference.