DframCy

DframCy is a light-weight utility module to integrate Pandas Dataframe to spaCy's linguistic annotation and training tasks. DframCy provides clean APIs to convert spaCy's linguistic annotations, Matcher and PhraseMatcher information to Pandas dataframe, also supports training and evaluation of NLP pipeline from CSV/XLXS/XLS without any changes to spaCy's underlying APIs.

Getting Started

DframCy can be easily installed. Just need to the following:

Requirements

• Python 3.6
• Pandas
• spaCy 2.2.0

Also need to download spaCy's language model:

python -m spacy download en_core_web_sm

For more information refer to: Models & Languages

Installation:

This package can be installed from PyPi by running:

pip install dframcy

To build from source:

git clone https://github.com/yash1994/dframcy.git
cd dframcy
python setup.py install

Usage

Linguistic Annotations

Get linguistic annotation in the dataframe. For linguistic annotations (dataframe column names) refer to spaCy's Token API document.

from dframcy import DframCy
dframcy = DframCy("en_core_web_sm")
doc = dframcy.nlp(u"Apple is looking at buying U.K. startup for $1 billion")

# default columns: ['id', 'text', 'start', 'end', 'pos', 'tag', 'dep', 'head', 'label'] 
annotation_dataframe = dframcy.to_dataframe(doc)

# can also pass columns names (spaCy's linguistic annotation attributes)
annotation_dataframe = dframcy.to_dataframe(doc, columns=["text", "lemma", "lower", "is_punct"])

# for separate entity dataframe
token_annotation_dataframe, entity_dataframe = dframcy.to_dataframe(doc, separate_entity_dframe=True) 

Rule-Based Matching

# Token-based Matching
from dframcy.matcher import DframCyMatcher, DframCyPhraseMatcher
dframcy_matcher = DframCyMatcher("en_core_web_sm")
pattern = [{"LOWER": "hello"}, {"IS_PUNCT": True}, {"LOWER": "world"}]
dframcy_matcher.add("HelloWorld", None, pattern)
doc = dframcy_matcher.nlp("Hello, world! Hello world!")
matches_dataframe = dframcy_matcher(doc)

# Phrase Matching
dframcy_phrase_matcher = DframCyPhraseMatcher("en_core_web_sm")
terms = [u"Barack Obama", u"Angela Merkel",u"Washington, D.C."]
patterns = [dframcy_phrase_matcher.get_nlp().make_doc(text) for text in terms]
dframcy_phrase_matcher.add("TerminologyList", None, *patterns)
doc = dframcy_phrase_matcher.nlp(u"German Chancellor Angela Merkel and US President Barack Obama "
                                u"converse in the Oval Office inside the White House in Washington, D.C.")
phrase_matches_dataframe = dframcy_phrase_matcher(doc)

Command Line Interface

Dframcy supports command line arguments for conversion of plain text file to linguistically annotated text in CSV/JSON format, training and evaluation of language models from CSV/XLS formatted training data. Training data example. CLI arguments for training and evaluation are exactly same as spaCy's CLI, only difference is the format of training data.

# convert
dframcy convert -i plain_text.txt -o annotations.csv -t CSV

# train
dframcy train -l en -o spacy_models -t train.csv -d test.csv

# evaluate
dframcy evaluate -m spacy_model/ -d test.csv