Pagination Construct Detection in Legacy Document Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for converting unstructured documents to structured formats, such as XML, often fail to efficiently and robustly identify and extract pagination constructs like headers and footers, leading to incorrect logical documents and noise in further processing.
Innovation Solution
A method and apparatus that detect pagination constructs by measuring text-box-to-text-box similarity based on textual content, position, and area, using a textual variability score to identify header and footer zones, and merging contiguous elements to enhance detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If headers and footers are not differentiated from body elements during conversion, then the conversion process is simpler, but the logical document becomes incorrect and processing noise is introduced
Solution Approach 1:
The patent segments the document into distinct elements by detecting pagination constructs (headers and footers) and separating them from body elements. This is achieved by identifying repeated text boxes across pages and marking them as pagination constructs, thereby creating a segmented structure that maintains logical correctness while preserving conversion simplicity.
Solution Approach 2:
The patent extracts headers and footers from the main document body by detecting their repeated occurrence across pages. The extraction process identifies text boxes that appear in multiple pages at consistent positions and separates them as pagination constructs, removing them from the body element flow to prevent processing noise.
2Difficulty of detecting and measuring
If neighboring page comparison techniques are used to identify headers/footers, then detection can be performed, but the method fails when headers/footers occur very few times in the document
Solution Approach 1:
The patent changes the detection parameter from frequency-based comparison to position-based similarity measurement. Instead of relying on how many times a header/footer appears, the system measures the similarity of text box positions and content across pages, allowing reliable detection even when pagination constructs occur only once or twice in the document.
3Device complexity
If page segmentation is discarded in logical representation, then the logical structure is simplified, but headers and footers cannot be properly preserved or differentiated
Solution Approach 1:
The patent performs preliminary detection of pagination constructs before final logical representation is created. By identifying headers and footers in advance through position and similarity analysis, the system can preserve this structural information in the logical representation without increasing overall complexity, as the detection is performed as part of the conversion process.
Data Source
AI summary
A method for identifying header/footer content of a document, in order to sequence text fragments comprising recognizable text blocks as derived from the document. The textual variability of lines comprised of text blocks, including the different kinds of text blocks within the line is analyzed for assessment of textual variability. Header/footer zones are defined by textual content having a low textual variability. An alternative embodiment identifies pagination constructs by comparing selected text-boxes for similarity and proximity and clustering the text boxes satisfying a predetermined similarity value, wherein the clustered text boxes are deemed to comprise pagination constructs.