What Distinguishes the Redfield NLP Solution from the KNIME NLP Approach?
- With the Redfield NLP Nodes you can perform NLP tasks fast and easy with no-code.
- The Redfield NLP Nodes offer the widest set of supported languages: currently 21 languages are released, and the number will grow.
- You can take advantage of the full set of functions, previously available for English language only.
- We offer a big set of community models ready for you to put into use (including HuggingFace repository).
- Easier text vectorization and full compatibility with the current set of Text Processing nodes
- Modern and accurate embedding
- BERT fine-tuned models can be easily used as regular Knime ML model for deployment
- New features that have never been available before
- Training of multi-label classification models
- Morphology functionality
- Support of dynamic tag sets for POS and NER
- We constantly work on the nodes maintenance and update them on a regular basis
Spacy extension
Text processing:
- Tokenization: split the document into words, this is an essential task for text processing
- Part-of-speech tagging: search meaningful parts of the sentence
- Named Entity Recognition: extract entities mentioned in the document
- Lemmatization: transform the words to their root form
- Morphology analysis: understand the relationships between subjects, objects and actions in the sentence
- Vectorization: represent words and documents as vectors so it can be used for machine learning
BERT extension
Process thousands of support conversations, customer reviews or social media posts. Feed in raw text – no text processing needed!
- Text classification: identify the type or sentiment of the documents
- Document classification: label documents according to its content
- Multi-Label document classification: assign different labels to a specific text according to its content
- BERT Embeddings and Similarity search: understand if documents are alike or different
