aniruddha27

doc = nlp(' Last year, I spoke about the Ujjwala programme , through which, I am happy to report, 50 million free liquid-gas connections have been provided so far')
png = visualise_spacy_tree.create_png(doc)
display(Image(png))

aniruddha27

# to extract initiatives using pattern matching
def all_schemes(text,check):
    
    schemes = []
    
    doc = nlp(text)
    
    # initiatives
    prog_list = ['programme','scheme',
                 'initiative','campaign',

aniruddha27

# to check if keyswords like 'programs','schemes', etc. present in sentences

def prog_sent(text):
    
    patterns = [r'\b(?i)'+'plan'+r'\b',
               r'\b(?i)'+'programme'+r'\b',
               r'\b(?i)'+'scheme'+r'\b',
               r'\b(?i)'+'campaign'+r'\b',
               r'\b(?i)'+'initiative'+r'\b',
               r'\b(?i)'+'conference'+r'\b',

aniruddha27

# function to find sentences containing PMs of India
def find_names(text):
    
    names = []
    
    # spacy doc
    doc = nlp(text)
    
    # pattern
    pattern = [{'LOWER':'prime'},

aniruddha27

import spacy
from spacy.matcher import Matcher 

from spacy import displacy 
import visualise_spacy_tree
from IPython.display import Image, display

# load english language model
nlp = spacy.load('en_core_web_sm',disable=['ner','textcat'])

aniruddha27

# split sentences
def sentences(text):
    # split sentences and questions
    text = re.split('[.?]', text)
    clean_sent = []
    for sent in text:
        clean_sent.append(sent)
    return clean_sent

# sentences

aniruddha27

# function to preprocess speech
def clean(text):
    
    # removing paragraph numbers
    text = re.sub('[0-9]+.\t','',str(text))
    # removing new line characters
    text = re.sub('\n ','',str(text))
    text = re.sub('\n',' ',str(text))
    # removing apostrophes
    text = re.sub("'s",'',str(text))

aniruddha27

for token in doc:
    # extract subject
    if (token.dep_=='nsubj'):
        print(token.text)
    # extract object
    elif (token.dep_=='dobj'):
        print(token.text)

aniruddha27

from spacy import displacy 
displacy.render(doc, style='dep',jupyter=True)

aniruddha27

text = "The children love cream biscuits"

# create spacy 
doc = nlp(text)

for token in doc:
    print(token.text,'->',token.pos_)