Text Expression Detection and Generation for Character Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating character-specific sentences in texts, such as scripts or novels, either require high costs and time for manual rewriting or generate inappropriate sentences due to the need for developing rules and machine learning models, leading to concerns about cost and accuracy in automated conversion methods.
Innovation Solution
An information processing device with a detection unit that uses a learning model to identify character-like expressions in text and a generation unit that creates alternative expressions based on detected features, allowing for user feedback to relearn the model and improve sentence generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual rewriting is used to generate character-specific sentences, then accuracy is improved, but time and cost increase significantly
Solution Approach 1:
The system implements feedback by detecting user reactions to generated different expressions and relearning the learning model accordingly. This allows the system to improve accuracy over time while maintaining automated operation, resolving the contradiction between high accuracy and low time/cost investment.
Solution Approach 2:
The system performs self-improvement by automatically relearning the learning model based on user feedback without requiring manual intervention for model training. This enables the system to maintain high accuracy while operating autonomously, reducing both time and cost.
2Loss of time
If automated sentence conversion based on rules or machine learning is used, then time and cost are reduced, but inappropriate sentences may be generated and development cost increases
Solution Approach 1:
The system detects user reactions to generated expressions and uses this feedback to relearn the learning model, continuously improving sentence quality while maintaining automated operation. This resolves the reliability issue by enabling the system to learn from actual usage and reduce inappropriate sentences over time.
Solution Approach 2:
The system dynamically adjusts the learning model parameters through relearning based on user feedback, allowing it to adapt to specific character traits and improve sentence appropriateness without requiring extensive initial rule development or model training.
3Extent of automation
If custom rules and machine learning models are developed for automated conversion, then sentence conversion capability is improved, but development cost increases
Solution Approach 1:
The system uses a general-purpose learning model that adapts to specific character traits through parameter adjustment via user feedback rather than requiring custom-designed rules or specialized models for each character. This reduces development complexity while maintaining high automation capability.
Solution Approach 2:
The learning model serves multiple functions: it detects character traits, generates appropriate expressions, and learns from user feedback across different characters and contexts. This universal approach eliminates the need for separate rule sets or models for each character, reducing overall system complexity.
Data Source
AI summary
An information processing device according to an embodiment includes: a detection unit (120) that detects an expression based on a feature amount extracted from a text, and character information including information of a character using a learning model learned in advance, the expression being included in the text and indicating the character likeness of the character; and a generation unit (140) that generates a different expression that is different from the expression and indicates the character likeness based on the expression detected by the detection unit and the character information, and presents the generated different expression, and the detection unit relearns the learning model according to a user's reaction to the different expression presented by the generation unit.