Handheld Text Input Disambiguation via External Language Object Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Generating text in handheld electronic devices, such as PDAs and cellular telephones, is complex due to the physical constraints of smaller keyboards, leading to ambiguous input and the need for complex disambiguation schemes, as existing solutions like adapting ten-digit keypads or shrinking traditional keyboards result in unclear character input.
Innovation Solution
A method and device that utilize lists of language objects, including words, abbreviations, and ideograms, to facilitate text generation by processing received text from external sources, such as e-mails, SMS, and MMS, to identify and integrate new language objects that meet specified characteristics, thereby expanding the available language objects and disambiguating inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If the keyboard is made smaller to reduce device size, then the device becomes more compact and portable, but the character input becomes ambiguous and more complex to process
Solution Approach 1:
The patent segments the character input process into multiple stages: initial ambiguous input from compact keyboard, followed by disambiguation using multiple candidate lists (generic words, application-specific words, learned words). This segmentation allows small keyboard physical constraints to be overcome through procedural division of the input task.
Solution Approach 2:
The system performs preliminary action by pre-computing and storing multiple candidate lists of possible words and phrases before input is needed. When a key sequence is entered, the disambiguation can immediately reference these pre-prepared lists rather than generating possibilities in real-time, expediting the resolution of ambiguous input.
2Measurement precision
If multiple lists are used to disambiguate key inputs, then character input accuracy improves, but the text generation process becomes more complex and time-consuming
Solution Approach 1:
The patent implements dynamics by making the disambiguation process adaptive rather than static. The system learns from user corrections and frequently used terms, dynamically updating the candidate lists and their priorities. This allows the complexity to be optimized over time based on actual usage patterns rather than relying on fixed, pre-configured lists.
Solution Approach 2:
The system performs self-service by automatically learning from user input patterns and corrections. When users select certain words from the candidate lists or correct misinterpretations, the system automatically updates its learned words list, reducing the need for manual configuration and improving accuracy without increasing operational complexity for the user.
3Adaptability or versatility
If the device learns new language objects from external text sources, then text generation adaptability improves, but the processing time and memory requirements increase
Solution Approach 1:
The patent applies preliminary action by pre-processing and storing language objects from external sources (emails, SMS, MMS) into structured candidate lists before they are needed for text generation. This advance preparation organizes the data in a readily accessible format, reducing the processing time required when actual text input is needed.
Solution Approach 2:
The system applies local quality by organizing learned language objects into context-specific lists (generic words, application-specific words, learned words) rather than maintaining a single monolithic database. This localized organization allows the system to quickly access only the relevant list for the current context, reducing search time and improving efficiency while maintaining high adaptability.
Data Source
AI summary
Incoming e-mails, instant messages, SMS, and MMS, are scanned for new language objects such as words, abbreviations, text shortcuts and, in appropriate languages, ideograms, that are placed in a list for use by a text input process of a handheld electronic device to facilitate the generation of text. Systems and methods consistent with the present invention may gather new language objects from sources of text external to the handheld electronic device, and may ignore new language objects that are not considered to be in a current language of the handheld electronic device.