Contextual Summarization Model Using Category Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text summarization models struggle with accurately summarizing text inputs containing multiple categories of content, often generating summaries based on irrelevant information due to the lack of contextual guidance.
Innovation Solution
A novel training process for summarization models that generates training samples combining textual content from different categories with corresponding labels and summaries, allowing the model to focus on a selected category for accurate summarization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a summarization model processes text containing multiple categories, then it can handle diverse content, but it generates summaries based on irrelevant information due to lack of contextual guidance
Solution Approach 1:
The patent segments the text processing by dividing the input text into multiple categories using category identifiers. Each category represents a distinct topic or subject area, allowing the model to process different portions of the text separately with appropriate contextual guidance for each segment.
Solution Approach 2:
The patent applies local quality by providing category-specific contextual guidance to different portions of the text. The system determines which category each text segment belongs to and applies相应的 summarization parameters and guidance tailored to that specific category, rather than using a uniform approach for all content.
2Quantity of substance
If the model generates summaries for all content, then it provides comprehensive coverage, but it includes irrelevant information from categories not of interest
Solution Approach 1:
The patent extracts only the relevant category information needed for summarization by using category identifiers to filter and select specific portions of the text. The system identifies and extracts text segments belonging to the target category while excluding irrelevant segments from other categories, providing focused summaries without unnecessary information.
Solution Approach 2:
The patent implements dynamic summarization by allowing the summarization behavior to change based on the selected category. The system can dynamically adjust which categories to summarize and which to exclude, providing flexible control over the summary content based on user needs and contextual requirements.
Data Source
AI summary
In one example, a system can receive a set of text samples and generate a set of summaries based on the set of text samples. The system can then generate a training dataset by iteratively executing a training-sample generation process. Each iteration can involve selecting multiple text samples from the set of text samples, combining the multiple text samples together into a training sample, determining a text category and a summary corresponding to a selected one of the multiple text samples, and including the text category and the summary in the training sample. After generating the training dataset, the system can use it to train a model. The trained model can then receive a target textual dataset and a target category as input, identify a portion of the target textual dataset corresponding to the target category, and generate a summarization of the portion of that target textual dataset.