Splitting AI#

Note

Contact us to enable this functionality.

Per Project users can train Splitting AIs.

To work with Splitting AIs, you should first check the relevant “Enable splitting” option in the settings for the Project. You will be prompted to do so when you visit the Splitting AIs list.

After training, users can activate one Splitting AI per Project. When a Splitting AI is active, newly uploaded multi-page Documents will be analyzed by the AI to determine whether the Document actually consists of multiple Documents. Users are then able to confirm or refine these suggestions in the Document Validation UI.

Note

The SmartView does not support splitting Documents, so the Document Validation UI will be enabled instead. You can switch back to the SmartView once the splitting suggestions have been reviewed.

For an overview of how Spliting AIs works together with Categorization AIs and Extraction AIs, see our architecture diagram.

Splitting AI Details#

Project#

The Project used to train the Splitting AI.

Status#

Statuses of an AI training.

  • “Queuing for training…”: The splitting AI is waiting in the queue for its training to be started.

  • “Data loading in progress…”: The training process has started and the Konfuzio server loads the training data into memory.

  • “AI training in progress…”: The training data is loaded into memory and the actual training takes place.

  • “AI evaluation in progress…”: “The splitting AI is trained and the evaluation of the trained splitting AI is conducted.”

  • “Training finished.”: The splitting AI is evaluated and can be used.

In case the splitting AI could not be trained it will have the status “Contact support”.

Description#

The description to Document the reason for training.

Version#

Incremented version per training

Created At#

Date and time when training was started.

Loading time (in seconds):#

Displays the average, minimum and maximum loading time across all runs of this AI.

Runtime (in seconds):#

Displays the average, minimum and maximum runtime across all runs of this AI. (If you add the loading and the runtime you get the overall time an AI run on a document has consumed.)

Train Splitting AI#

The training process is 100 % automated, so the only setup users need to add a short description. The short description will help to relate the intention behind any change in the project to the quality of the Splitting AI.

In order to train a Splitting AI, documents from at least two different Categories need to be in the training set.

Retrain Splitting AI#

If you have new Documents uploaded to your Project you can train a new version of your Splitting AI.

  1. Add those to the Status: Training documents

  2. Train Splitting AI, see above.

  3. As you use the same documents with Status: Test documents but increased the number of documents with Status: Training documents the AI quality should improve.

Splitting AI Actions#

Evaluate Splitting AIs#

If you change the documents assigned to the status test dataset you can evaluate old Splitting AI models. This is helpful to evaluate different Splitting AIs on the current test dataset.

Activate Splitting AI#

A handy option to choose which Splitting AI is currently active in the project.