Newer
Older
import random # library for generating random numbers
import nltk # library for working with human language
from nltk.tokenize import word_tokenize # A function from NLTK for breaking words down
from nltk import ConditionalFreqDist # A class from NLTK for representing the conditional frequency distribution of a set.
nltk.download('punkt') # models used by word_tokenize to tokenize words.
def probs_model(text): # takes text as input
words = word_tokenize(text) # split words
bigrams = list(nltk.bigrams(words)) # pairs of consecutive words
cond_freq_dist = ConditionalFreqDist(bigrams) # conditional frequency of the 2 words
return cond_freq_dist
# probs_model is the representation of the conditional probs.
def generate_next_word(model, initial_word): # takes in the representation of the conditional probs and the initial word
next_words = model[initial_word] # randomly selects a next word based on the conditional probs
if not next_words:
return None # in case there are no next words
next_word = random.choices(list(next_words))[0]
return next_word
if __name__ == '__main__':
input_text = "In the sweet town of Candyland, there lived a marshmallow named Mallow. Mallow had a unique passion ? a love for Alan Turing's work on computers and artificial intelligence. Instead of bouncing with other candies, Mallow spent its days reading Turing's papers and dreaming of marshmallow-powered machines. Mallow's friends couldn't quite understand its fascination, but they embraced Mallow's uniqueness. One day, Mallow surprised everyone by creating a tiny marshmallow computer that could solve candy puzzles. The town marveled at Mallow's ingenuity, and Mallow's love for Turing's work became a source of inspiration for Candyland. And so, Mallow, the marshmallow with a Turing twist, continued to blend sweetness with technology, making Candyland a tastier and smarter place."
# Reads the content of republic.txt and stores it in the variable input_text.
model = probs_model(input_text)
# Selects any word and generates the next word based on conditional probabilities
initial_word = "Alan"
generated_next_word = generate_next_word(model, initial_word)
# Print the input text and the next word.
print("Input text:", input_text)
print(f"Given seed word: '{initial_word}', Generated next word: {generated_next_word}")