Overview

The Symanto Brain extension is a powerful tool that enables users to generate AI models in over 50 languages within the KNIME workspace. It is powered by a comprehensive, in-house pre-trained Transformer model, advanced NLP techniques and machine learning algorithms, that allow you to create text classification models that can categorise novel, unlabeled classes without the need for labelled training data. The extension overcomes the limitations of pre-built models and elevates text analytics capabilities, thereby enhancing the value of your insights, all within the KNIME platform.

Installation

1. To install the Symanto Brain extension, first, make sure that you have the latest version of KNIME installed on your system.

2. Open the KNIME website, and then the Community Hub.

3. Search for Symanto Brain and select it.

4. Simply drag & drop the extension from KNIME Hub into your running KNIME Analytics Platform.

Configuration

Once you have installed the extension and the KNIME Analytics Platform has automatically restarted, you will be able to see Symanto Brain in the Nodes Repository panel. If you expand it, you can access the Symanto Brain nodes.

Authentication Node

In order to use the Symanto Brain extension and its nodes, you need an API key:

1. Drag and drop the Authentication node.

2. Right-click and select Configure..

3. Add your API key.

   1. If you don’t have one, you can navigate to our Developer portal by clicking on the Get API key button. Alternatively, you can access the portal with the following link: developers.symanto.com. 

   2. Please sign up on the developer portal or send an email to support@symanto.com to obtain your access token. 

   3. Once your account is verified, please go to the free trial section to get your free key, which you can afterwards find in your profile section.

4. Click on the Apply button. If all settings are valid, the node status will change to configured ⚪🟡⚪.

5. Right-click and execute the node.

Open

Zero-shot

1. Drag and drop the Zero-shot node.

2. Right-click and select Configure.. A dialogue window will open:

Open

1. Add the five key elements required to configure your model: 

• Patterns - known as label descriptors, they provide context to the model and allow the semantic matching between the analyzed text and the different labels, e.g. This text is {} or This person talks about {}

• Language - the language of your text. You can select it from the drop-down menu.

• Text column - the field which will be considered as text in your data file.

d. Model - the embedding model that will be used.

1. Click on the

Apply button. If all settings are valid, the node status will change to configured ⚪🟡⚪.

1. Right-click and execute the node.

1. When the status is changed to Executed

⚪⚪🟢 , the Zero-shot node will calculate and produce a table. You can see the table from the Node Monitor at the bottom.

A sample workflow:

Technical Details

This extension is compatible with KNIME version 4.x and above.

The extension is designed to work with a wide range of data sources and languages.

Symanto Brain is deployed on a highly efficient and scalable cloud infrastructure provided with reliable GPU infrastructure, which enables high efficiency in both model creation and subsequent consumption. 

Support

You can join the KNIME Community and discover solutions, support, and inspiration from other KNIME users at https://forum.knime.com/?src=knimeapp.

In case you are experiencing any issues using the Symanto Brain extension, please contact our support team at support@symanto.com.