src.dackar.RCA.ner.hybrid_ner.classifier
========================================

.. py:module:: src.dackar.RCA.ner.hybrid_ner.classifier


Classes
-------

.. autoapisummary::

   src.dackar.RCA.ner.hybrid_ner.classifier.TrainingExample
   src.dackar.RCA.ner.hybrid_ner.classifier.SpanClassifier


Module Contents
---------------

.. py:class:: TrainingExample

   Minimal supervised example for the span classifier.

   - doc_text: full document text
   - start/end: span offsets in doc_text
   - label: gold label (must exist in schema.label_to_group)


   .. py:attribute:: doc_text
      :type:  str


   .. py:attribute:: start
      :type:  int


   .. py:attribute:: end
      :type:  int


   .. py:attribute:: label
      :type:  str


.. py:class:: SpanClassifier(context_window_chars = 50, top_k = 3, min_prob = 0.25)

   Lightweight ML span classifier placeholder using scikit-learn.

   Purpose:
     - Propose label hypotheses for candidates that are currently unlabeled ("defer").

   Usage:
     - clf = SpanClassifier()
     - clf.fit(examples, schema)
     - pipeline = HybridNERPipeline(..., classifier=clf)

   If scikit-learn is not available, this becomes a no-op (safe).


   .. py:attribute:: context_window_chars
      :value: 50



   .. py:attribute:: top_k
      :value: 3



   .. py:attribute:: min_prob
      :value: 0.25



   .. py:attribute:: _is_trained
      :value: False



   .. py:attribute:: _vectorizer
      :value: None



   .. py:attribute:: _model
      :value: None



   .. py:attribute:: _sklearn_ok
      :value: False



   .. py:method:: fit(examples, schema)


   .. py:method:: predict(doc, candidates, schema)


   .. py:method:: _featurize(text, start, end)