VectorSpaceModel Carrot2

Question

Is it possible to get the vector space model after you have clustered your documents?

I see in the documentation, it is possible to create your own Vector Space Model with:

public VectorSpaceModelContext(PreprocessingContext preprocessingContext)

And the prepocessingcontext would be:

PreprocessingContext(LanguageModel languageModel, List<Document> documents, String query)

With my list of documents, but that would be before I have clustered my documents.

I want the vector space model for the clusters.

Last resort would be to create it myself...

Answer 1

The only way would currently be modifying the source code of the algorithm to expose the VSM as one of the output attributes. To do this, you'd need:

1. Define the Output attribute for your VSM model (an example for the Lingo algorithm)

2. Save the created VSM model to the attribute (an example for the Lingo algorithm)