Text Input Suggestion Ranking via Utility Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text input methods on computing devices often fail to provide timely suggestions for character strings, as users may spend more time selecting from lists of suggested strings than typing the remaining characters, especially if the suggested strings are not the most likely to occur in the given language context.
Innovation Solution
A computing device determines a utility score for each character string based on its probability of occurrence, time saved by selecting it, and time expended to analyze and select it, then outputs the most useful character strings for display, prioritizing those that accelerate user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the computing device provides a list of suggested character strings for display, then the user can select complete character strings to accelerate text entry, but the user may spend more time analyzing and selecting from the list than typing the remaining characters
Solution Approach 1:
The patent changes the parameter of suggestion ranking by introducing a scoring mechanism that weighs multiple factors: probability of the character string occurring in the given language context, time saved by selecting it, and time expended to analyze and select it. This parameter transformation resolves the contradiction by dynamically adjusting which suggestions are presented based on real-time analysis of user input patterns and language probabilities.
Solution Approach 2:
The patent applies local quality by providing different levels of suggestion detail based on the specific language context and input state. Instead of uniformly presenting all possible character strings, the system locally adapts the suggestion list to show only the most relevant options for the current language and input situation, reducing analysis time while maintaining productivity benefits.
2Productivity
If the computing device displays suggested character strings, then users can complete text entry faster, but users may waste time on suggestions that are not likely to occur in the given language context
Solution Approach 1:
The patent transforms the suggestion generation process by introducing a probability-based scoring system that evaluates language context. The system calculates the probability of each character string occurring in the given language context and uses this to weight the suggestion ranking. This parameter change ensures that computational resources are focused on generating and displaying only the most likely character strings, eliminating waste on unlikely suggestions while maintaining high text entry efficiency.
3Reliability
If the computing device provides multiple suggested character strings, then users have more options for accurate text entry, but the complexity of determining and presenting the right suggestions increases
Solution Approach 1:
The patent applies segmentation by breaking down the complex suggestion determination process into distinct analytical components: probability analysis of language context, time saved calculation, and time expended estimation. Each component processes information independently and contributes to the final scoring. This segmentation reduces overall complexity by making the suggestion determination process more modular and manageable, while still providing accurate and reliable text entry suggestions.
Data Source
AI summary
A computing device receives an indication of user input that corresponds to a series of one or more characters. The computing device determines, based at least in part on the series of one or more characters, a plurality of character strings. The computing device determines a score for each respective character string based at least in part on a probability in a particular language of the respective character string occurring, an estimate of time to be saved by a user if the user were to select the respective character string, and an estimate of time to be expended by the user to analyze the respective character string. The computing device outputs at least a portion of the plurality of character strings. The portion of the plurality of character strings is determined based at least in part on the score for each respective character string from the plurality of character strings.