MLP Support Teacher for Grading

 # Step 1: Import libraries

import pandas as pd

from sklearn.feature_extraction.text import TfidfVectorizer

from sklearn.tree import DecisionTreeClassifier

from sklearn.model_selection import train_test_split

from sklearn import tree

import matplotlib.pyplot as plt

# Step 2: Create a simple dataset (Question + Subject + Difficulty)

data = {

'question': [

"Explain photosynthesis",

"Define gravity",

"What is chemical bonding",

"Describe kinetic energy",

"Explain mitosis",

"What is quantum mechanics",

"Describe water cycle",

"Explain Ohm's Law",

"What are atoms made of",

"Explain the nervous system"

],

'subject': [

"Biology", "Physics", "Chemistry", "Physics",

"Biology", "Physics", "Geography", "Physics",

"Chemistry", "Biology"

],

'difficulty': [

"Easy", "Easy", "Medium", "Medium",

"Easy", "Hard", "Easy", "Medium",

"Medium", "Hard"

]

}

# Step 3: Convert into DataFrame

df = pd.DataFrame(data)

# Step 4: Text Preprocessing (TF-IDF Vectorizer)

vectorizer = TfidfVectorizer()

X = vectorizer.fit_transform(df['question'])

# Step 5: Encode Subject Labels

subject_labels = df['subject']

difficulty_labels = df['difficulty']

# Step 6: Train two Decision Trees: one for subject, one for difficulty

X_train, X_test, y_train_subj, y_test_subj = train_test_split(X, subject_labels, test_size=0.2, random_state=42)

_, _, y_train_diff, y_test_diff = train_test_split(X, difficulty_labels, test_size=0.2, random_state=42)

model_subject = DecisionTreeClassifier()

model_difficulty = DecisionTreeClassifier()

model_subject.fit(X_train, y_train_subj)

model_difficulty.fit(X_train, y_train_diff)

# Step 7: Function for AI Assistant

def ai_teaching_assistant(query):

query_vec = vectorizer.transform([query])

predicted_subject = model_subject.predict(query_vec)[0]

predicted_difficulty = model_difficulty.predict(query_vec)[0]

print(f"Predicted Subject: {predicted_subject}")

print(f"Predicted Difficulty: {predicted_difficulty}")

# Give a simple response

if predicted_subject == "Biology":

print("Here’s a resource: https://www.khanacademy.org/science/biology")

elif predicted_subject == "Physics":

print("Check out this: https://www.physicsclassroom.com/")

elif predicted_subject == "Chemistry":

print("Learn more at: https://www.chemguide.co.uk/")

elif predicted_subject == "Geography":

print("Explore: https://www.nationalgeographic.org/")

else:

print("No resource found.")

# Step 8: Test the Assistant

print("\n=== Example Run ===")

user_query = input("Ask your question: ")

ai_teaching_assistant(user_query)

# Optional: Visualize one of the trees (Subject)

plt.figure(figsize=(15,10))

tree.plot_tree(model_subject, filled=True, feature_names=vectorizer.get_feature_names_out(), class_names=model_subject.classes_)

plt.title("Decision Tree for Subject Classification")

plt.show()