IME Platform Context-Aware Candidate Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional input method editors (IMEs) lack the ability to provide diverse and context-aware input candidates tailored to specific application scenarios, language modes, and user preferences, limiting their functionality and user experience.
Innovation Solution
An advanced IME that selects and provides text, rich, and informative candidates based on user input, scenario context, and language mode, utilizing machine learning to adjust scores and rankings, and allowing users to associate and launch multiple applications to enhance input options, with a client-server architecture for efficient candidate generation and display.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single source is used to identify input candidates in conventional IME, then the system complexity is low, but the diversity and relevance of input candidates are limited
Solution Approach 1:
The patent segments the candidate generation function into multiple independent applications (e.g., dictionary application, translation application, search application). Each application operates independently to provide candidates for specific scenarios, allowing the system to maintain low complexity for each component while achieving high diversity through the combination of multiple specialized applications.
Solution Approach 2:
The IME is designed to universally support multiple types of applications and scenarios. The system can adaptively select from different applications based on the current input scenario, making the IME versatile across various contexts (chatting, document editing, gaming) while maintaining manageable complexity through scenario-based routing.
2Loss of information
If conventional IME provides only text candidates, then the system is simple, but the user experience and information quality are limited
Solution Approach 1:
The patent applies local quality by providing different types of candidates (text, rich media, informative content) based on the specific scenario and user needs. For example, in a translation scenario, the system provides both text translations and rich media content; in a dictionary scenario, it provides text definitions and informative explanations. This localized enhancement of candidate quality improves information delivery without requiring the system to handle all candidate types uniformly.
3Adaptability or versatility
If the IME does not support application extensions, then the system is simple, but the functionality and adaptability are limited
Solution Approach 1:
The patent introduces an intermediary mechanism (the application selection module) that mediates between the user input and multiple candidate applications. This intermediary receives the input scenario, selects the appropriate application based on predefined rules and machine learning scores, and routes the request accordingly. This mediator approach enables functional extensibility through multiple applications while keeping the core IME architecture relatively simple.
Solution Approach 2:
The system dynamically selects which applications to activate based on the current input scenario and context. The application selection is not static but adapts in real-time based on scenario analysis, machine learning score rankings, and user preferences. This dynamic approach allows the system to extend functionality only when and where needed, maintaining simplicity in scenarios where extensions are not required.
4Measurement precision
If machine learning techniques are used to adjust application scores, then the candidate ranking accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent applies partial action by using machine learning techniques selectively rather than comprehensively. The system uses machine learning to adjust scores for applications that benefit from contextual understanding, while relying on simpler rule-based selection for straightforward scenarios. This partial application of machine learning improves ranking accuracy for complex cases without incurring the full computational cost of applying ML to every candidate selection.
Data Source
AI summary
An input method editor (IME) provides a distributed platform architecture that enables associating multiple applications with the IME to provide extended functionalities. The presentations of the applications, such as skins, may be different from each other and that of the IME. The applications may be represented in a manifest file that is human-readable and editable. The IME collects multiple parameters relating to a user input into a host application including a query input by the user and a scenario of the host application, and selects one or more applications to provide candidates based on a score or ranking of the applications under the collected multiple parameters. Machine-learning may be used to improve the score or ranking. The candidates may include text candidates, rich candidates, and informative candidates.