Implement a BIO parser module

This package will take BIO files as input. To make any further manipulations easier, we should convert that into a more common format like a Python Class. We can take inspiration from spacy's Doc (see their doc).

added P2 label

assigned to @yschneider

mentioned in merge request !5 (closed)

Some of the metrics use sequences of Named Entity Blocks for computation. Therefore, it would make the implementation job less costly if this Python class has a method to obtain a list of the Named Entity Blocks from the BIO file (or from the Python object). Example:

From: B-Person John I-Person Meadows O was O a O famous O bodybuilder O for O the B-Organization IFBB

Obtain: ne_blocks = [("Person", "John Meadows), ("Organization", "IFBB")]

Implemented in ner/bio-parser!2 (merged) and added as a submodule in !7 (merged)

closed