Word Vector Apply

This node applies a Word Vector Model to Documents or Strings, meaning that every word contained in the text (tokenized by whitespace) will be replaced by the corresponding Word Vector if the word is available in the Word Vector dictionary. This will result in a list of Word Vectors. If specified, the node calculates the mean of these vectors resulting in a single representation of the input.

Options

Text Column

The column containing the text where the Word Vector Model should be applied to. The text will we tokenized by whitespace. Depending on the type of the input the node will use the following as input:

• String: the String will be directly used
• Document: The Document Full Text will be appended to the Document Title separated by a whitespace.
  E.g. a Document with Title: 'Doc A' and Full Text: 'This is a test'
  will result in the following text: 'Doc A This is a test'

Calculate Mean Vector?

Whether to calculate the mean of the word vectors.

Views

This node has no views

Workflows

• 04_Word_Embedding_ClaxKNIME Hub
• 07_Simple_Document_Classification_Using_Word_VectorsKNIME Hub
• KNIME_textanalysis_group_project_PAKNIME Hub
• KNIME_textanalysis_group_project_PA_finalKNIME Hub
• pos_lstmKNIME Forum
• Show all 7 workflows

Links

• No links available

Developers