Text Detection Using Contextual Features for Low-Quality Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing text quality detection methods fail to accurately identify low-quality texts, especially in different scenarios, and struggle to recognize newly emerging low-quality expressions, as they only consider the text itself without contextual information.
Innovation Solution
A text detection method that determines first and second attribute features of a text and associated elements, along with their association relationships, and inputs these features into a trained network model to improve detection accuracy, considering contextual information such as author and reader behavior patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If only the text itself is considered for quality detection, then the detection process is simple, but the detection accuracy deteriorates because the same text may express different meanings in different scenarios
Solution Approach 1:
The patent transitions from one-dimensional text analysis to multi-dimensional analysis by incorporating scenario information, author features, and reader features as additional dimensions. This allows the system to distinguish the same text in different contexts (e.g., medical scenarios vs. daily life scenarios), resolving the contradiction between simple processing and accurate detection.
2Device complexity
If traditional text classification models are used, then the model structure is simple, but the ability to recognize newly emerging low-quality expressions deteriorates
Solution Approach 1:
The patent performs preliminary actions by pre-collecting and storing scenario information, author features, and reader features before text detection. This preparatory data collection enables the model to quickly adapt to new low-quality expressions by leveraging pre-processed contextual information, improving versatility without proportionally increasing structural complexity.
Solution Approach 2:
The patent creates a composite detection system that combines multiple types of data (text content, scenario information, author features, reader features) into a unified detection framework. This composite approach enables the model to recognize diverse and emerging low-quality expressions by synthesizing information from multiple sources.
3Measurement precision
If contextual information from associated elements is incorporated, then detection accuracy improves, but the data processing complexity increases
Solution Approach 1:
The patent segments the complex detection task into distinct modules: text feature extraction, scenario information processing, author feature processing, and reader feature processing. Each module handles specific aspects independently, then results are integrated for final detection. This segmentation improves detection precision while managing processing complexity through modular organization.
Data Source
AI summary
Provided are a text detection method and apparatus, an electronic device and a storage medium. The method includes: determining a first attribute feature of a to-be-detected text and a second attribute feature of elements each having an association relationship with the to-be-detected text; inputting the first attribute feature, the second attribute feature, association relationships between the to-be-detected text and the elements, and association relationships between the elements into a trained network model to obtain a detection result of the to-be-detected text. Such technical solution improves a detection accuracy of a low-quality text.