Pipelines integration

[nicolay-r]

It might be useful to take a look onto document processing framework in terms of variety pipelines that allows you to extract a particular type of information.
Such information might be as follows: NER, contexts, context-level attitudes, document level attitudes [many more in a potential future]

• Integration for text processing
• Enhance pipeline results organization (automatically assign src value once pipeline element accomplised and provide the related result).
• Enhance ppl.run() by adopting src-based parameters as a separate input_data argument.

The latter might be gathered into pipelines as follows:

# implement pipeline
ppl = Pipeline(
               ParseDocument(
                   SentenceSplitter, NER, ...
               ), 
               ExtractContexts(...),
               AnnotateContexts(...),
               ConvertToDocumentLevel(...))
                      
# call
ppl()