Multilingual Name Recognition via Language-Specific Grammatical Rules
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing software systems struggle to recognize names in computer-generated documents across multiple languages due to differences in grammatical rules and language contexts, making it difficult to provide helpful actions to users.
Innovation Solution
A method and system that utilize a name recognizer application, which breaks text into words using a word breaker application if necessary, applies language-specific grammatical rules to isolate names, and compares them against user contact lists or larger databases to provide helpful actions such as adding, deleting, or sending emails to recognized names.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If software attempts to recognize names in multilingual documents, then helpful actions can be provided to users, but recognition accuracy decreases due to diverse grammatical rules and language contexts
Solution Approach 1:
The patent segments the name recognition task by creating separate recognizer components for different languages (e.g., English recognizer, Thai recognizer, Vietnamese recognizer). Each recognizer is specialized for its language's grammatical rules, allowing the system to maintain high accuracy for each language while supporting multiple languages overall. The system selects the appropriate recognizer based on the detected language of the input text.
Solution Approach 2:
The patent applies local quality by making each language recognizer have specialized properties tailored to its specific language's grammatical rules and naming conventions. For example, the Thai recognizer is configured with Thai-specific rules for detecting names, while the English recognizer uses English-specific rules. This allows each component to operate with high precision for its designated language.
2Ease of operation
If the system supports multiple user interface languages and creates documents in various languages, then usability improves for international users, but the complexity of recognizing names across different grammatical rules increases
Solution Approach 1:
The patent implements a universal architecture where a single host application can create and process documents in multiple languages. The system includes a language detection mechanism that universally handles any input language, and a set of language-specific recognizers that can be selected based on the detected language. This universal design allows the system to support many languages without requiring separate applications for each language.
Solution Approach 2:
The patent introduces language-specific recognizer components as intermediaries between the host application and the name recognition task. When the host application needs to recognize names, it detects the language of the input text and passes it to the appropriate recognizer intermediary, which then applies the correct grammatical rules. This intermediary layer manages the complexity by isolating language-specific logic from the main application.
3Extent of automation
If names are recognized by comparing against contact lists and databases, then helpful actions can be provided, but the difficulty of distinguishing names from other words increases
Solution Approach 1:
The patent applies preliminary action by first detecting the language of the input text before attempting to recognize names. The system performs language detection as a preliminary step, then selects the appropriate language-specific recognizer that has pre-configured grammatical rules and naming patterns for that language. This preliminary language identification makes the subsequent name recognition easier and more accurate.
Solution Approach 2:
The system uses feedback by comparing recognized name candidates against contact lists and databases, then using the results of this comparison to refine and confirm the name recognition. If a candidate name matches entries in the contact list or database, this feedback confirms the recognition. The system iteratively refines its recognition based on this feedback from external data sources.
Data Source
AI summary
Methods and systems are provided for recognizing names entered into a computer-generated document in the context of a plurality of spoken languages and for providing helpful actions to users upon the recognition of the entered names.