A submission is created when a set of files are submitted into Hyperscience together. The system then interprets each image as a page and matches each page to a live layout in the system. Pages are grouped into a document based on the matched layout.
The Submissions page contains the following tabs, which display details of all submissions in the application:
Submissions
Documents
Cases
No Layout Found
Potential Layouts
Accepted file types
We support the following file types:
JPEG
PNG
TIFF
HEIC
HTML
PDF
Editable PDFs are not supported.
ZIP
XPS
MSG (files and their attachments)
[v38.1.8 and later] EML (files and their attachments)
You can disable the processing of files attached to EML files. For more information, see Processing EML files' attachments.
Note that if the ZIP file is password-protected or the images are encrypted, an error will appear.
Not all of the file types are supported in their native formats. For example, to process HTML submissions, the system uses images of the HTML files rather than the HTML files themselves.
Not all of these file types can be used to train Classification models. For more information, see Model Management.
Support for other file types is possible but with limitations. For more information, contact your Hyperscience representative.
Supported languages
Hyperscience supports the automation of submissions written in any of our supported languages.
For more information, see Supported Languages.
Submission statuses
A submission goes through multiple steps in Hyperscience to complete extraction and completing a step changes the status of a document.
Below is a simplified Document Processing Graph that shows the machine and manual tasks a submission can go through. It includes both Structured and Semi-structured document statuses.
For more information on the manual tasks shown in this diagram, see the article What is Supervision?
Processing Graph
Status aggregation
If pages and documents in a submission have different statuses, the earliest status in the Document Processing Graph will be shown. The same is true for pages in a document.