Home
Production
Descriptions of OCR processing, TEI transformation, and metadata creation
Page Files
Explains the procedures we use to get the best quality OCR of each page.
Draw Boxes
Manually creating text recognition boxes improves accuracy
Table Areas
Creating table areas improves recognition

Production
Descriptions of OCR processing, TEI transformation, and metadata creation
- Organization
  How to keep hundreds of thousands of files organized.
- Page Files
  Explains the procedures we use to get the best quality OCR of each page.
  - AFR Interface
    Learn about the main elements of the program interface
  - Create a Page-Inventory File
    Create a page-inventory file.
  - Create an Image Collection
    Organize image files for scanning.
  - Create an OCR-Project
    How to create and manage an OCR-Project.
  - Settings
    Recommended settings for all options in ABBYY FineReader
  - Draw Boxes
    Manually creating text recognition boxes improves accuracy
    - Text Areas
      How to draw text areas
    - Reorder Areas
      How to order areas correctly for text output.
    - Modify Areas
      Areas must be drawn in a basic box shape, but they can be altered to subtract or add to part of the box.
    - Footnote Areas
      Footnotes always output as the last areas on the page.
    - Table Areas
      Creating table areas improves recognition
      - Edit tables
        Table columns and rows can be adjusted before recognition.
    - Headers
      How to treat two-column entry titles
    - Illustrations
      When drawing boxes, leave out illustrations and their captions.
    - Catchwords
      Leave out the extra word at the bottom of the page in eb03
    - Author Initials
      Always include author indicators at the end of articles
    - Pages without Text
      When to include pages without text in the output
    - Special Cases
      Situations requiring a note in the page-inventory file.
  - Page Recognition
    Excellent page recognition depends on preparing pages properly.
  - Save and Output
    How to output your OCR results.
  - TEI-XML
    This introduction to Oxygen XML Editor shows you how to navigate the interface and perform standard procedures on the Encyclopedia files.
- Entry Files
  Procedures for converting single pages into Encyclopedia entries.
- Master Files
  After creating entry files, we run an automated metadata generation process to add index terms to each entry.

Table Areas

Creating table areas improves recognition

All tabular matter should be identified as a table, not text. This includes material without lines or rules, but where the text aligns in more than one column.

At this time, tables should be identified as such but they do not need to be corrected during the OCR process. Let ABBYY FineReader do its best to automatically recognize the material. We will return to tables later in the production process.

For tables, use the Draw Table Area tool. Complex pages can have multiple tables, requiring a combination of multiple text and table areas. This takes time and managing the numbering of your boxes becomes critical.
Figure 1. Page with both text and table areas.
All tabular matter should be identified as a table, even if it does not have rules around it. This is critical to success in the OCR process.
When tabular data is treated as text instead of as a table, it results in output that might look acceptable in AFR and even in the docx file. However, Word formats it as a list, and it will not convert properly into TEI, requiring extra work to adapt it to a table format. The work goes faster if such material is formatted as a table at the initial OCR stage.

Figure 2. Tabular matter without rules around it.

The first instance of tabular matter is correctly identified as a table, but the second is not and needs to be corrected by creating a four-column table with six rows. See Edit tables.