This script find similarity between submissions.

similarity.py

import pandas as pd
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity

# Load CSV
df = pd.read_csv("submissions.csv")

# Use the 'rationale' column
rationales = df['rationale'].fillna('')

# Convert text to TF-IDF vectors
vectorizer = TfidfVectorizer(stop_words='english')
tfidf_matrix = vectorizer.fit_transform(rationales)

# Compute cosine similarity between all pairs
similarity_matrix = cosine_similarity(tfidf_matrix)

# Threshold for "almost same" (0.8 means 80% similar)
threshold = 0.8

# Find similar submissions
similar_pairs = []
for i in range(len(df)):
    for j in range(i + 1, len(df)):
        if similarity_matrix[i, j] >= threshold:
            similar_pairs.append((df.iloc[i]['user_name'], df.iloc[j]['user_name'], similarity_matrix[i, j]))

print(f"Found {len(similar_pairs)} pairs of students with similar answers:")
for pair in similar_pairs:
    print(f"Candidates: {pair[0]} and {pair[1]} - Similarity: {pair[2]:.2f}")