Hierarchical Text Classification for Accurate PDF Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data extraction techniques from PDF documents are error-prone and time-consuming, failing to accurately extract data in a correct hierarchical format.
Innovation Solution
A method and system that utilize a combination of predefined rules and machine learning to classify text data in a document, determining positional and text-characteristic information for line regions, and selecting hierarchy classifications based on probability values to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual extraction is used to achieve accurate hierarchical data extraction, then data accuracy is improved, but time consumption and labor cost increase
Solution Approach 1:
The patent replaces manual mechanical extraction processes with an automated system combining optical character recognition (OCR), machine learning classification, and rule-based processing. The system automatically performs hierarchical classification of extracted text data, substituting human labor with computational processes that achieve both accuracy and efficiency.
Solution Approach 2:
The patent introduces an intermediary classification system between raw text extraction and final data output. This intermediary layer uses machine learning models and predefined rules to hierarchically organize extracted text, enabling automated processing while maintaining the accuracy that would otherwise require manual intervention.
2Productivity
If automated extraction techniques are used to reduce time consumption, then productivity is improved, but data extraction accuracy deteriorates
Solution Approach 1:
The patent segments the extraction process into distinct hierarchical levels (first hierarchy classification and second hierarchy classification). Each segment handles specific aspects of classification, with the first level providing broad categorization and the second level refining the classification. This segmentation enables automated processing while maintaining accuracy through multi-stage verification.
Solution Approach 2:
The patent incorporates feedback mechanisms where extracted text is processed through multiple classification stages, with each stage refining the output of the previous stage. The system uses feedback from the first hierarchy classification to inform the second hierarchy classification, enabling automated processing to achieve accuracy comparable to manual extraction.
3Speed
If simple text extraction is used to achieve fast processing, then speed is improved, but hierarchical structure preservation deteriorates
Solution Approach 1:
The patent performs preliminary hierarchical classification during the extraction process itself, rather than as a separate post-processing step. By integrating classification into the extraction workflow, the system preserves hierarchical structure while maintaining fast processing speeds, as the classification is performed concurrently with text extraction.
Solution Approach 2:
The patent adds a hierarchical dimension to the extracted text data by implementing multi-level classification. The system transforms flat extracted text into a structured hierarchical format through first and second hierarchy classifications, preserving the document's structural relationships while maintaining processing efficiency.
Data Source
AI summary
A method and system for classifying text data in a document based on hierarchy classification is disclosed. A plurality of line regions comprising text data in the document are determined. Positional information and text-characteristic information for each of the plurality of the line regions is determined. For each of the plurality of line regions, a first hierarchy classification from the plurality of hierarchy classifications based on a plurality of predefined rules is determined and a second hierarchy classification from the plurality of hierarchy classifications and a respective probability value based on a machine learning technique is determined. Each of the plurality of line regions are classified based on the first hierarchy classification or the second hierarchy classification. The second hierarchy classification is selected in case the respective probability value of the second hierarchy classification is greater than equal to a predefined threshold.