This Python script provides a utility to compute the cosine similarity between two text sentences. This script is useful for applications like plagiarism detection, semantic search, or text data preprocessing where understanding the degree of similarity between texts is crucial.

prompt-compare.py

"""
This Python script provides a utility to compute the cosine similarity between two text sentences using the TF-IDF 
(Term Frequency-Inverse Document Frequency) vectorization approach. 

Key Components:
1. Import Statements: The script begins by importing necessary modules:
   - TfidfVectorizer from sklearn.feature_extraction.text for converting text data into a matrix of TF-IDF features.
   - cosine_similarity from sklearn.metrics.pairwise to compute the similarity between two vectors in the TF-IDF space.
   - sys for accessing command-line arguments.

2. Function get_sentences(): This function retrieves the sentences either from command-line arguments or interactive 
user input if no arguments are provided. It ensures the program can work flexibly in different usage contexts.

3. Function calculate_similarity(): Accepts two sentences and computes their similarity. It uses the TF-IDF vectorizer 
to transform the sentences into vectors and then calculates the cosine similarity between these vectors. The similarity 
is also converted into a percentage format for easier interpretation.

4. Function main(): Orchestrates the flow by calling get_sentences() to retrieve the sentences, computes their similarity 
using calculate_similarity(), and then prints the results.

5. Execution Check: The script checks if it's run as the main program and calls the main() function accordingly, ensuring 
it doesn't run when imported as a module.

This script is useful for applications like plagiarism detection, semantic search, or text data preprocessing where 
understanding the degree of similarity between texts is crucial.
"""

from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
import sys

def get_sentences():
    # Check if sufficient arguments are provided
    if len(sys.argv) == 3:
        sentence1, sentence2 = sys.argv[1], sys.argv[2]
    else:
        # Prompt user for input if arguments are not provided
        sentence1 = input("Please enter the first sentence: ")
        sentence2 = input("Please enter the second sentence: ")

    return sentence1, sentence2

def calculate_similarity(sentence1=None, sentence2=None):
    if sentence1 is None or sentence2 is None:
        print("Error: You have to pass two sentences for calculating similarity", file=sys.stderr)
        return None, None
    
    # Vectorize the sentences
    vectorizer = TfidfVectorizer()
    vectors = vectorizer.fit_transform([sentence1, sentence2])
    
    # Compute cosine similarity
    cosine_sim = cosine_similarity(vectors[0:1], vectors[1:2])[0][0]
    
    # Convert to percentage
    percentage_similarity = 50 * (cosine_sim + 1)  # Scale from 0% to 100%
    
    return cosine_sim, percentage_similarity

def main():
    sentence1, sentence2 = get_sentences()
    print(f"Sentence 1: {sentence1}")
    print(f"Sentence 2: {sentence2}")
    similarity_score, similarity_percentage = calculate_similarity(sentence1, sentence2)
    print(f"Cosine Similarity Score: {similarity_score}")
    print(f"Similarity Percentage: {similarity_percentage}%")

if __name__ == "__main__":
    main()