Input Support Code Output Using Cluster Similarity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing input support apparatuses using automatic classification functions in a sequential processing method often reduce usability due to rapid changes in displayed classification codes that do not conform to user intuition, leading to reduced value and efficiency in input operations.
Innovation Solution
An input support apparatus that divides input data into tokens, estimates classification codes using a classification model, constructs a cluster space, and adjusts the output method based on cluster similarity to stabilize and emphasize estimation codes, ensuring they align with user intuition and improve usability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the estimation code is updated every time a character is input, then the responsiveness of the system is improved, but the stability of the displayed code deteriorates
Solution Approach 1:
The patent implements dynamic output control by adjusting the output timing and frequency of estimation codes based on input length thresholds and confidence levels. The system transitions from static frequent output to dynamic adaptive output, updating codes selectively based on whether input reaches certain lengths or confidence exceeds thresholds, thus balancing responsiveness with stability
Solution Approach 2:
The patent changes the parameter of output frequency from constant (every character input) to variable based on input length and confidence level. By introducing length thresholds and confidence thresholds as controllable parameters, the system adapts its output behavior to match user needs and input characteristics, resolving the contradiction between rapid updates and stability
2Measurement precision
If the classification code is estimated from hundreds of candidates, then the accuracy of classification is improved, but the time required for estimation increases
Solution Approach 1:
The patent applies partial action by initially presenting a limited set of high-confidence estimation codes (e.g., top 3 candidates) rather than processing all hundreds of candidates immediately. When users need more options or the system confidence is sufficiently high, it expands to show more candidates or finalizes the classification, thus achieving accurate classification without always incurring the full time cost of evaluating all candidates
Solution Approach 2:
The patent performs preliminary filtering and ranking of classification candidates before presenting them to users. The estimation model pre-processes hundreds of candidates by calculating confidence scores and sorting them, so that when output is triggered, only the most relevant candidates are displayed. This preliminary action reduces the perceived estimation time while maintaining high accuracy
3Adaptability or versatility
If the estimation code changes frequently, then the system adapts to user input dynamically, but the usability deteriorates due to codes not conforming to user intuition
Solution Approach 1:
The patent implements feedback mechanisms where the system monitors user interactions (such as code selection, rejection, or manual input) and uses this feedback to adjust future estimation behavior. By learning from user choices and confidence level patterns, the system refines its adaptation strategy to better align with user intuition while maintaining dynamic responsiveness to input changes
Solution Approach 2:
The patent introduces dynamic control of code presentation based on confidence levels and input progression. Instead of uniformly changing codes with every input, the system dynamically adjusts when to update, when to maintain the current code, and how many candidates to display, creating a more usable adaptive behavior that respects user intuition while remaining responsive
Data Source
AI summary
An input support apparatus includes processing circuitry configured to detect a division position in input data and divides the input data into a plurality of division units, acquire an estimation code by estimating a classification code of the input data by using a classification model that receives an input of data and estimates the classification code in a classification system, classify the classification code into a plurality of clusters using explanatory information of the classification code and constructs a cluster space by embedded expression, calculate a cluster similarity between the input data and the cluster by using the cluster space, and adjust a method of outputting the estimation code based on the cluster similarity.