Text Embedding refers to the process of computing a numerical representation of a piece of text (often a sentence), that can then be used as a feature vector for Machine Learning.
Text Embeddings are computed using large-scale embedding models that generate vectors that are close for related pieces of text.
Sentence embedding is a native feature handling option for text features in Visual ML. Please see Text variables for more information. With this method, you can benefit for high quality extraction from text features without any specific configuration or work
In addition to the native feature, you can also use the “Sentence Embedding” plugin. This plugin provides a recipe that allows you to retrieve the sentence embeddings directly as a vector column. This can be used for further customized processing in code. This can also be used for similarity search
This feature is only available in English.
This capability is provided by the “sentence-embedding” plugin, which you need to install. Please see Installing plugins.
This plugin is Not supported
Please see our plugin page for more information