Template definition and field placement

This guidance covers the following:

What is a template?

A template is a blank form used to match form submissions within the SS&C Chorus Document Automation interface. All form submissions are sorted in the system for a good template match. A successful form-to-template match will allow fields to guide the location of the data extraction on a filled in form. Templates must be structured in order for fields to be found in the same spatial region as the submitted filled in form.

Template terms and definitions

  • Template – A set of sheets that constitutes all template pages and its versions.

  • Template page – A single page of a template (relevant for multi-page templates).

  • Version – A sheet representing one possible variation of a given template page.

  • Sheet – An image of a file represented as a single sheet of a template page, version, and/or a filled in form.

Template rules and behavior

  • A single template page can be comprised of multiple sheets as versions of itself.

  • All sheets in the template page share the same schema fields.

  • Fields can be activated and deactivated on each sheet, depending on whether the fields appear on the page.

File size requirements

In the examples below, each template is a US Letter page measuring 8.5 inches in width by 11 inches in height. The green boxes on the page represent how large a field can be drawn on the template. The current maximum field (shred) size (height x width) is 250,000 pixels (e.g. 500 x 500, 2500 x 100, or 1250 x 200). Any field larger than the maximum field size will be returned as --impossible--.

There are a couple of factors that help you achieve optimal document recognition and data extraction:

  1. The template and expected page size and resolution should ideally be the same.

    1. To capture data successfully using Document Automation's READ, it is important that the page size of your submitted files is as close to, or ideally the same, as the sheet you’ve uploaded into your template.

      When you submit image files for digitization, the document recognition engine re-sizes each incoming page to the width and height of your template. This will affect the ability to form a cleaner comparison of the pages and for the data to return an accurate transcription. Ideally the templates and submitted file pages are 8.5 x 11 at 150 pixels per inch (PPI).

    2. The page size of the file can be as small as half of the recommended 8.5 x 11 and as large as three to four times larger (178,956,970 pixels), but the expected submitted page size must be the same.

  2. To have the ideal resolution of 150 PPI.

    1. The model was primarily trained on shreds drawn on template sheets at a resolution of 150 PPI. We recommend the minimum resolution of the file to be 150 PPI and a maximum of 200 PPI to achieve optimal data quality. The file page size resolution is typically described in PPI.

      The higher the resolution the more pixels are used to represent a page. For example, a US Letter page at 150 PPI that has a resolution of 1275 pixels wide by 1650 pixels high, whereas a higher definition image of the same page at 600 PPI is 5100 x 6600 pixels.

    2. A template’s image resolution (PPI) affects how large a field can be drawn on a template and how shreds are processed using that field. The smaller (less PPI) your template has, the larger area a field can cover. The larger (more PPI) your template has, the smaller area a field will cover. However, a template created using a page that is very low definition, like 100 PPI, may result in reduced accuracy.

Form structure best practice

Templates are used to visually match incoming forms for digitization. Templates should be structured or semi-structured to ensure that data extraction will be consistent across all pages. When forms are submitted, the sorting engine will match the forms to existing templates and the predefined data fields are then shredded (cropped) for digitization. A clean, blank, and structured form is ideal to upload as a template.

Types of form structure

Document Automation's alignment algorithm uses markers to identify consistent structural differences between forms and defined templates in order to match filled in forms to blank templates.

Structured forms have fields in clearly marked, delineated, and fixed locations. These fields stay consistent across the form, though may vary across form versions.

  • Structured – Structured forms have fields in clearly marked, delineated, and fixed locations. These fields stay consistent across the form, though may vary across form versions.

  • Semi-structured – Semi-structured forms have select fields in fixed locations, but may have data in tables with a variable number of lines. Data can be captured from the template areas with fixed locations.

  • Unstructured – Unstructured forms are highly variable and often contain a large field for unconstrained text, such as diary pages, books, lab slips, or newspapers. Unstructured forms are not supported at this time.

Characteristics of clear form structure

  • Every field should have a clearly separated area, large enough to capture complete answers. For example:

  • Fields with open spaced lines with the field label to the left of it allows uninterrupted capture.

  • The form text should not be in close proximity to the space provided for the data. This ensures that only data of interest is captured by our machine. Thus, the best practice to draw fields and capture data is to avoid capturing the form text.

  • Add a page number and/or form code on each page. This may help in the case of pages being lost or shuffled. If you have multiple-page templates or versions of your form, name the file with the form code and/or the page number on each page.

  • The data on the form should be found in the same location as its corresponding template. The machine expects the templates (blank forms) and filled in forms to be exact visual matches, even slight visual or text differences can alter your results.

Template creation

Templates are created and managed in the Templates dashboard page. All created templates are listed with the following template information:

  • Template name.

  • Template status.

  • The number of pages.

  • The time stamp that the template was last modified.

The default view when entering the Templates page is ordered by the most recently edited active and draft templates. On your Templates dashboard, the listed templates can be searched through in the search bar, sorted (by oldest first, most recent, A-Z, or Z-A), and filtered (by template status).

Templates dashboard actions

In the drop-down menu next to Edit Template, users are able to organize pages, change template status, duplicate templates, rename templates, and adjust template settings.

The template status will affect certain actions, as described below, and can be changed in the drop-down menu to the right of Edit Template:

Status

Description

You can edit the template, use it to process batches, and view data associated with it.

You can view data associated with the template.

You can edit the template and view data associated with it.

Create a template

  1. Click the orange Create a Template button on the top-right hand corner.

  2. Enter a unique template name in the modal.

  3. Click the orange Upload Pages button. Browse files from the computer, or drag and drop files onto the webpage. The files must be in PNG, GIF, JPG, or PDF format.

    Each page of the form does not have to be saved as a separate file. Document Automation arranges batch files alphanumerically by file name.

Template naming best practice

Document Automation templates should be created based on context from the form. Template names may also be streamlined to only have the necessary information required. For example, naming the template for "2020 Application for Deferred Annuity" as "App_Deferred_Annuity_2020".

Best practices for naming templates are as follows:

  • Use a unique template name for each template.

  • Avoid using spaces and utilize underscores “_” where possible.

  • Use the sheet name to specify different version details.

    • The default sheet name will be designated at the filename upon upload.

    • The sheet name can be edited at any time.

  • Utilize common naming conventions across the forms. Two templates of the same type can share the prefix of the template name to indicate similar template types. For example, two HIPAA Authorization forms that are designated as two distinct template types could be defined as “HIPAA_Auth_2012” and “HIPAA_Auth_2019”.

  • (Optional) If the form contains a form code, it is recommended to use it in the template name.

Template organization actions

Template pages and versions may be organized immediately after they are uploaded. This page can also be accessed by clicking Organize Pages in the template drop-down menu. The user will be led to a full overview of the template pages in order from left to right. The template pages can be ordered by clicking on the template page and dragging it before or after another template page.

Users can view and edit the template page by clicking the Edit button on the top-right hand corner or double-clicking into the specific file image. You can also view the template by clicking on the image file and then clicking the eye icon on the top-right hand corner.

  • Rename a sheet – Click the image file of interest then the three dots icon button. Click Rename in the drop-down menu.

  • Replace a sheet – Click the image file of interest then on the three dots icon button. Click Replace image in the drop-down menu.

  • Delete sheet – Click the image file(s) and then the trash can icon. A pop-up will ask you to confirm the action. This action is only available if the template has not been used to digitize forms.

Versions of the template

Client forms are typically provided as a set, with corresponding version types. These versions do not necessarily require their own template in the system as they can be grouped together as a version of one template. Only unique versions will be uploaded as a template page.

A version is any individual sheet, representing one possible variation of a given template page. A version should be created if a sheet has slight visual differences such as changes to form text, spacing, font, or additional fields, to ensure complete and accurate data capture. Uploading the correct version is required for sorting configuration in System Set-up and can be determined through Version Discovery. This is the process of determining how many templates and versions to build to ensure incoming sheets are sorted correctly.

For more information, see Version Discovery.

Version rules and behaviors

  • Versions allow data extraction of visually similar (but not duplicate) sheets to be digitized.

  • Once fields are placed on one version of the template, it will appear in the exact location across all versions of that template page. Those fields will need to be moved, resized, or made inactive.

Creating versions

If there are multiple versions in a template page, be sure to upload the files in page order. You can organize the versions and pages in the template preview page.

  1. To add versions to your template, click on the Upload Pages tab on the top-right hand corner. Once your pages have been uploaded, it will appear after template pages that have already been uploaded.

  2. Click the sheets to be grouped into versions and drag it over the template page or version.

    A single click will highlight the sheet you've clicked on.

  3. Once you've successfully grouped the versions, confirm the action in the modal by clicking the blue Group button. The versions will appear grouped under the template page in the template preview page. If the template page has versions, the versions preview page can be accessed to rename, replace, or delete a sheet. By double-clicking into a version, the user will be led to the template edit page. In the template edit page, the template pages will be found on the left and its individual versions in the drop-down menu.

Field behavior across versions

All sheets in the template page share the same schema fields, this means a new field does not have to be created every time it appears on a particular version. Once placed on a version of the template page, it will appear in the exact location across all versions. Any following action to move the field will not apply to the other versions of the template page.

Use the Active and Inactive buttons in the field list to activate fields that are present, or inactivate fields that are missing from that version. For more information, see Field placement on versions.

Field placement

  1. On the Templates dashboard, click on the Edit Template button to enter the template edit page.

    The template image appears on the left with a panel to the right.

  2. On the right panel, click the plus sign (+) on the top right corner of the panel, or Add Fields in the center of the panel.

    The user is presented with a list of fields and its data type previously defined in the Field Library.

  3. Drag the intended field and drop it on the template image in the desired location. Once dropped onto the template image, a green outlined box will appear over the location where the original blue box was dropped. Click and drag the green borders to re-size the field. The green box should cover where a user would fill in that field on the page, without overlapping additional and irrelevant text and characters. On the right panel, the field list will show the user whether the schema field has been placed, indicated by a green check mark.

    1. By double-clicking the field in the panel on the right or on the image to the left, the user will be able to view the field details. Each field can only be placed once per template.

  4. Repeat this process for all schema fields required for placement on the template. A complete template will look like the following image, with each active field listed on the right panel corresponding to its field on the template image.

Best practices of field placement

The way that fields are drawn on the template can affect READ's ability to accurately transcribe data that appears on the form. It should be placed exactly in the area of capture and sized as tightly as possible to only capture the necessary information.

For multiple choice fields on the form, fields should be drawn as tightly as possible around the checkbox or choice bubble to lead to better data quality. Avoid text and answer fields (blank fields to write in additional values) when it is possible. Please refer to the best practices of field placement for multiple choice fields in the following image:

Schema field behavior

  • Across different templates – If “prop_ins_first_name” appears on Template 1 and Template 2, it can be used on both Templates 1 and Template 2.

  • Across the same template – “child_1_first_name” can only be used once in a single page of Template 1, it can't be used again on another page within Template 1.

Field placement on versions

For every field placed on any version of the template page, the same field will automatically be made active across all the other versions of that page in the exact location. The fields in the other versions in the template page will likely need to be moved, re-sized, or made inactive. The best practice is to add the fields on the first version of the template and then edit that field on the subsequent versions of that template page.

For example, “claimant_gender” may be made active on Version A, but inactive on Version B. To make the a field inactive on a specific version of a template, click on the Active button next to the field's name in the Template definition page.

Inactive multiple choice values

Multiple choice values listed in the Multiple Choice Selection in the Field Library can be inactivated and activated as needed on a per-version basis.

To inactivate the multiple choice value, double-click into a field in the template edit page to view the field details. Click on the Active button to inactive that multiple choice value. The multiple choice value can be activated again by clicking on the Inactive button.

Please be sure that the Multiple Choice Selection has been set up with the values listed from left to right, then top to bottom. For more information about multiple choice field setup, see this article.

Template settings

The order in which fields have been placed on the template will be the order that fields are presented in the case file by default. Since that order may not follow the visual flow of reading, fields can be and are encouraged to be ordered by their position in the template from left to right, then top to bottom.

  1. Navigate to the Templates dashboard. For the template you would like to edit, click in the template’s drop-down menu, then click Settings.

  2. On the Template Settings page, click Field Ordering.

  3. On the Automatically order option, click on Position to update the field ordering to the position of fields as it appears on the template. An Autosaved message will appear in green above the field list once your changes have been made.

Template status after digitization

Templates will have limited behavior once it has been digitized against a form. After digitization, templates become “Safe Editable” Templates. Please be careful since these are not noted or indicated with a label in the Templates dashboard.

Acceptable actions after digitization:

  • Field management

    • Move fields

    • Resize fields

    • Add fields

    • Remove fields

    • Inactivate fields

  • Edit fields in the field library

  • Add versions

Actions unavailable after digitization:

  • Removing existing versions (this requires duplication of the template)