Chording Keyboard Mapping Table Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing chording keyboard optimization methods fail to effectively separate perceptual and motor scores, measure transitions accurately, and account for common letter sequences, leading to suboptimal keyboard mappings that do not cater to all user types, including beginners and fast typists, and have not resulted in a widely adopted or commercially successful one-handed keyboard.
Innovation Solution
A method that uses a two-dimensional array to measure psychomotor costs of transitions, allows any chord to transition directly to another without releasing keys, and incorporates a text corpus to optimize chord maps, considering sequences and commands, with a permutation-computation-selection cycle to converge on an optimal mapping table.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pure timing is used to disambiguate chords, then chord recognition accuracy is improved for fast typists, but beginners are excluded because their slow typing speed causes extraneous chord recognition
Solution Approach 1:
The system dynamically adjusts the disambiguation method based on the user's typing speed and skill level. For beginners with slower typing speeds, the system uses probabilistic models that account for learning curves and allow more time for chord recognition. For advanced users with faster typing speeds, the system switches to pure timing-based disambiguation to maximize recognition accuracy and typing speed.
Solution Approach 2:
The system changes the disambiguation parameters based on user performance metrics. Typing speed, accuracy, and error rates are monitored to dynamically adjust the threshold for chord recognition. This allows the system to adapt to different user groups, accommodating both beginners who need more lenient recognition criteria and fast typists who benefit from strict timing-based disambiguation.
2Device complexity
If a keyboard requires all keys to be released between chords, then chord transition complexity is reduced, but motor activity increases by more than twice compared to direct transition keyboards
Solution Approach 1:
The system prepares for the next chord transition by maintaining a buffer of pre-processed chord states. Instead of requiring complete key release, the system anticipates upcoming chords and pre-positions the keyboard state, allowing smoother transitions without the need to fully release all keys. This reduces motor activity while maintaining transition clarity.
Solution Approach 2:
The system enables continuous chord transitions without interrupting the typing flow. By allowing direct transition from one chord to another without requiring complete key release, the system maintains continuous useful action (typing) rather than introducing idle periods where keys must be released and repressed. This significantly reduces motor activity while preserving transition precision through software-based disambiguation.
3Ease of operation
If optimization is based on distance from resting state of finger, then single-finger keyboard efficiency is improved, but it is ineffective when all fingers operate only a single key
Solution Approach 1:
The optimization system is designed to be universal, supporting both single-finger and multi-finger operation modes. The chord mapping table can be optimized for different finger configurations and operation styles. The system adapts its optimization criteria based on the detected usage pattern, making it equally effective whether users operate with one finger or multiple fingers, thus achieving both single-finger efficiency and multi-finger versatility.
4Device complexity
If one-dimensional cost table is used for chord entry cost, then storage simplicity is maintained, but it cannot capture perceptual and motor score relationships
Solution Approach 1:
The system transitions from a one-dimensional cost table to a two-dimensional cost surface that captures both perceptual and motor score dimensions. Instead of storing simple scalar costs, the system maintains a matrix where rows represent perceptual scores and columns represent motor scores, allowing comprehensive analysis of chord entry costs. This dimensional expansion enables more precise optimization while the use of efficient data structures keeps storage requirements manageable.
Data Source
AI summary
A method of creating a chording keyboard mapping table is described. An exercise is created that is performed by users of a keyboard for which a mapping table is desired. The psychomotor performance of the users' transitions between chords is recorded—as distinct from performance relating to chords. A first trial keyboard mapping table (from chords to symbols) is created, and the effectiveness of that table is measured by applying it to a text corpus and measuring the total psychomotor cost to enter that corpus with that table. The table is then permuted, and the cost recomputed. Trial tables are repeatedly permuted, with one or more preferred tables being used as the basis for the next iteration of permutations. This continues until a terminating event, such as minimal improvement, is reached. Variations as embodiments are described. Software to implement the method is claimed.