Document Reading Order Detection Using Layout-Aware Semantic Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining the reading order of text elements in documents, particularly in rich-text documents, is challenging due to their flexible layout and arrangement, leading to inaccuracies in subsequent processing tasks.
Innovation Solution
A computer-implemented method that determines the reading order of text elements by generating semantic feature representations based on both text sequence and layout information, using machine learning techniques to accurately characterize the spatial layout of text elements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fixed strategies are used to determine reading order, then the method is simple and fast, but the accuracy deteriorates in rich-text documents with flexible layouts
Solution Approach 1:
The patent transforms the reading order detection problem by changing parameters: it converts spatial layout information into semantic features and uses attention mechanisms to dynamically weight different text elements. This allows the system to adapt to various document layouts without requiring complex rule-based systems, resolving the contradiction between accuracy and complexity.
Solution Approach 2:
The patent replaces traditional mechanical rule-based reading order determination with a neural network-based semantic analysis system. The model automatically learns reading patterns from layout information and text semantics, substituting rigid mechanical strategies with adaptive intelligent processing that achieves higher accuracy without proportional increases in system complexity.
2Measurement precision
If layout information is incorporated to improve reading order detection, then accuracy improves, but processing complexity increases
Solution Approach 1:
The patent introduces semantic feature representations as an intermediary between raw layout information and reading order determination. The layout information is first transformed into semantic features that capture the essential spatial relationships, which then feed into the reading order model. This intermediary step simplifies the processing by converting complex spatial data into meaningful semantic representations.
3Measurement precision
If semantic feature representations are generated based on both text sequence and layout information, then reading order accuracy improves, but computational resources increase
Solution Approach 1:
The patent applies partial action by focusing the semantic feature generation on the most relevant aspects of layout information and text sequences. Rather than processing all possible features equally, the model selectively attends to key spatial relationships and text elements that most influence reading order, reducing unnecessary computational overhead while maintaining accuracy.
Data Source
AI summary
According to embodiments of the present disclosure, there is provided a solution for reading order detection in a document. In the solution, a computer-implemented method includes: determining a text sequence and layout information presented in a document, the text sequence comprising a plurality of text elements, the layout information indicating a spatial layout of the plurality of text elements in the document; generating a plurality of semantic feature representations corresponding to the plurality of text elements based at least on the text sequence and the layout information; and determining a reading order of the plurality of text elements in the document based on the plurality of semantic feature representations. According to the solution, the introduction of the layout information can better characterize a spatial layout manner of the text elements in a specific document, thereby determining the reading order more effectively and accurately.