calculator.py

# code to calculate Rouge precision and recall for various texts, by taking the two text files 
# that have the original summaries and the new ones.

# Inputs:

# 1) File containing original summaries
# 2) File containing new summaries
# 3) n-gram model to use (1, 2, 3 ...) (Should be less than the total number of words in any paragraph)

# Output

# Baseline_n_gram.txt containing tab delimited Precision and Recall values for each pair.
import numpy as np

def rouge_metrics(system_list,reference_list):
    reference_word_count = len(reference_list)
    system_word_count = len(system_list)
    if (system_word_count == 0) or (reference_word_count == 0):
        rouge_recall = 0
        rouge_precision = 0
    else:
        rouge_recall = len(intersection(system_list,reference_list))*1.0/reference_word_count
        rouge_precision = len(intersection(system_list,reference_list))*1.0/system_word_count
    return rouge_recall, rouge_precision

def intersection(system_lst, ref_lst):
    intersection_lst = [value for value in system_lst if value in ref_lst]
    return intersection_lst

def create_ngrams(text_list,n=2):
	iterations = len(text_list)-n
	ngrams = []
	gram = []
	for i in range(iterations+1):
		gram = text_list[i:n+i]
		ngrams.append(gram)

	return ngrams


original_summaries_file_path = "y_data_train.txt"
new_summaries_file_path = "baseline.txt"
ngram = 2

with open(original_summaries_file_path, "r") as f:
    original = f.read()

with open(new_summaries_file_path, "r") as f:
    new = f.read()

rrecall = []
rprecision = []

for row_original, row_new in zip(original.split("\n"), new.split("\n")):

	with open("baseline_" + str(ngram) + "_gram.txt", "a") as f:

		system_list = row_original.split(" ")
		reference_list = row_new.split(" ")

		system_2grams = create_ngrams(system_list,ngram)
		reference_2grams = create_ngrams(reference_list,ngram)

		rouge_2_recall, rouge_2_precision = rouge_metrics(system_2grams,reference_2grams)

		rrecall.append(rouge_2_recall)
		rprecision.append(rouge_2_precision)

		line = str(rouge_2_recall) + "\t" + str(rouge_2_precision) + "\n"
		f.write(line)


rrecall = np.mean(rrecall)
rprecision = np.mean(rprecision)
f_score = 2*rprecision*rrecall/(rprecision + rrecall)

print ("Recall:", rrecall)
print ("Precision:", rprecision)
print ("FScore:", f_score)