Narrative Text Rewriting via Contextual Edit Operations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine-based solutions struggle to generate a coherent ending for a narrative text when an antecedent condition changes, often requiring significant modifications to the original story and facing challenges in controlling language models to avoid overediting.
Innovation Solution
A method and apparatus for rewriting narrative text that involves determining changes to sentences, performing edit operations on text parts based on inconsistencies with the target context, and replacing original text parts with edited versions to produce a rewritten narrative text, while ensuring contextual coherence and minimizing edits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional machine-based solutions are used to generate a coherent ending when an antecedent condition changes, then the narrative text can be rewritten, but a large number of edits are required and contextual coherence is difficult to maintain
Solution Approach 1:
The patent segments the narrative text into sentences and identifies specific sentences that contain the changed antecedent condition. By working at the sentence level rather than rewriting the entire text, the system maintains contextual coherence while minimizing the number of edits required. The segmentation allows focused editing only where necessary.
Solution Approach 2:
The patent applies local quality by performing edit operations only on specific text parts that are inconsistent with the target context, rather than uniformly editing the entire narrative. The system identifies text parts that need modification and applies edits locally, preserving the original text where it remains coherent and consistent with the changed condition.
2Adaptability or versatility
If significant modifications are made to the original story to accommodate changed conditions, then the narrative can adapt to new conditions, but the original story structure and coherence are compromised
Solution Approach 1:
The patent applies partial action by performing only the necessary edit operations to accommodate the changed antecedent condition, rather than making comprehensive modifications to the entire story. The system identifies the minimal set of text parts that require editing to achieve adaptability, thus preserving the original story structure while still adapting to new conditions.
Solution Approach 2:
The patent changes parameters of specific text parts (such as replacing words, modifying phrases, or adjusting sentence structures) to accommodate the changed antecedent condition. By modifying parameters of individual text parts rather than the overall story structure, the system achieves adaptability while maintaining structural stability.
3Reliability
If language models are used to generate coherent endings, then narrative coherence can be improved, but the models tend to overedit and make excessive changes to the original text
Solution Approach 1:
The patent implements feedback by evaluating the consistency between text parts and the target context, and using this evaluation to guide edit operations. The system continuously monitors whether edits are improving coherence without causing excessive changes, adjusting its editing strategy based on the measured consistency between the modified text and the changed antecedent condition.
Solution Approach 2:
The patent performs preliminary action by first identifying sentences and text parts that require editing before making any modifications. The system pre-processes the narrative text to locate the minimal set of changes needed, thereby preventing overediting while ensuring that necessary edits are made to maintain coherence with the changed conditions.
Data Source
AI summary
According to embodiments of the present disclosure, a method and apparatus for rewriting a narrative text, a device, and a medium are provided. The method includes determining a change to a sentence in a narrative text. An initial context of the sentence before the change is different from a target context of a changed sentence. The method further includes performing, based on inconsistency between a text part after the sentence in the narrative text and the target context, at least one edit operation on the text part to generate at least one edited version of the text part. The method further includes replacing the text part with an edited version in the at least one edited version to obtain a rewritten narrative text. In this way, the narrative text can be rewritten with a small number of edits while ensuring contextual coherence.