Skip to content

NLP Core Knowledge Model

The NLP Core Knowledge Model (display name: NLP Core EN v#) performs these analyses on English texts:

  • Deep linguistic analysis, which, in turn, comprises:
    • Text subdivision
    • Part-of-speech tagging
    • Morphological analysis
    • Lemmatization
    • Syntactic analysis
    • Semantic analysis
  • Keyphrase extraction
  • Named entity recognition
  • Relation extraction
  • Sentiment analysis

These are the same analyses that any model carries out as a preliminary step to extract the features on which the ML algorithm or the symbolic rules base their predictions, but in this case the model is limited to such analyses. Therefore, if all that is needed is one or more of the above analyses, this is the model to use.
It is also true that the output of these analyses can be obtained from any model, in addition to the predictions of categories and extracted information, but, by carrying out this activity only, this Knowledge Model takes less memory and is faster than predictive model.

Output structure

The model output has the same structure as any other model and is affected by the functional properties of the workflow block, but the categories and the extractions arrays are always empty.

This is the mapping between the analyses and the output keys containing their outcome:

AnalysisCorresponding output sections
Keyphrase extractionmainLemmas
Morphological analysistokens
Named entity recognitionentities
Part-of-speech taggingtokens
Relation extractionrelations
Semantic analysistokens
Sentiment analysissentiment
Syntactic analysistokens
Text subdivisionparagraphs