Shared Corpus Word Prediction for Reduced Keyboards
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Handheld computing devices with reduced keyboards face challenges in designing intuitive, fast, and powerful text entry user interfaces due to increased opportunities for user error and slower data entry.
Innovation Solution
Implementing a system that uses a shared corpus to supplement word prediction, completion, and correction by searching a remote site with user input, where the client device transmits text input to a server for search results to enhance prediction and correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Weight of moving object
If a reduced keyboard is used in handheld devices, then device portability is improved, but text entry accuracy deteriorates
Solution Approach 1:
A shared corpus stored on a server acts as an intermediary between the user's partial input and the complete word suggestions. The corpus contains pre-stored text data that the server queries to provide accurate word completions and corrections, compensating for the limited keyboard input capability.
Solution Approach 2:
The corpus is pre-populated with extensive text data before use. This preliminary preparation of the corpus allows the system to quickly provide accurate word suggestions without requiring real-time computation, thereby maintaining text entry accuracy despite the reduced keyboard.
2Weight of moving object
If a reduced keyboard is used in handheld devices, then device portability is improved, but data entry speed deteriorates
Solution Approach 1:
The system provides immediate feedback by displaying multiple word suggestions based on the user's current input. This feedback loop allows users to quickly select from suggested words rather than typing complete words, significantly accelerating data entry speed on reduced keyboards.
Solution Approach 2:
The corpus is pre-populated with extensive text data before use. This preliminary preparation of the corpus allows the system to quickly provide accurate word suggestions without requiring real-time computation, thereby maintaining text entry accuracy despite the reduced keyboard.
3Device complexity
If word prediction is performed locally without a shared corpus, then system simplicity is improved, but prediction accuracy deteriorates
Solution Approach 1:
A shared corpus stored on a server acts as an intermediary between the user's partial input and the complete word suggestions. The corpus contains pre-stored text data that the server queries to provide accurate word completions and corrections, compensating for the limited keyboard input capability.
Solution Approach 2:
The system transitions from local-only word prediction to a distributed architecture involving client devices and remote servers. This dimensional expansion allows the system to leverage remote computational resources and large-scale corpora without increasing local device complexity.
4Measurement precision
If a shared corpus is searched remotely, then prediction accuracy is improved, but system complexity deteriorates
Solution Approach 1:
The system transitions from local-only word prediction to a distributed architecture involving client devices and remote servers. This dimensional expansion allows the system to leverage remote computational resources and large-scale corpora without increasing local device complexity.
Solution Approach 2:
The complex corpus storage and processing functionality is extracted from the client device and placed on remote servers. This extraction allows the client device to remain simple while still accessing the power of a comprehensive shared corpus through networked queries.
Data Source
AI summary
Searching a shared corpus is used to supplement word prediction, completion, and/or correction of text entry. A user input device at a client device receives user entry of text input comprising a string of symbols. The client device wirelessly transmits instructions to a remote site to conduct a search of a corpus using the string as a contiguous search term. From the remote site, the client device receives results of the search, including multiple sets of one or more words, each set occurring in the corpus immediately after the search term. The client device uses the received sets in word prediction, completion, and/or correction.