Create app.py

app.py

@@ -0,0 +1,86 @@
+import os
+import spacy
+from sklearn.feature_extraction.text import TfidfVectorizer
+from sklearn.metrics.pairwise import cosine_similarity
+
+# Load legal data - Cases
+cases_directory = '/kaggle/input/legalai/Object_casedocs/'
+cases_texts = []
+
+for file_name in os.listdir(cases_directory):
+    file_path = os.path.join(cases_directory, file_name)
+    with open(file_path, 'r') as file:
+        content = file.read()
+        cases_texts.append(content)
+
+        # Load legal data - Statutes
+statutes_directory = '/kaggle/input/legalai/Object_statutes/'
+statutes_texts = {}
+for file_name in os.listdir(statutes_directory):
+    file_path = os.path.join(statutes_directory, file_name)
+    with open(file_path, 'r') as file:
+        statute_content = file.read()
+        # Use the file name (statute name) as the key and content as the value in the dictionary
+        statutes_texts[file_name] = statute_content
+
+# Preprocess and vectorize text for cases
+nlp = spacy.load("en_core_web_sm")
+processed_cases_texts = [" ".join([token.lemma_ for token in nlp(text) if not token.is_stop]) for text in cases_texts]
+vectorizer_cases = TfidfVectorizer()
+tfidf_matrix_cases = vectorizer_cases.fit_transform(processed_cases_texts)
+
+# Preprocess and vectorize text for statutes
+processed_statutes_texts = [" ".join([token.lemma_ for token in nlp(text) if not token.is_stop]) for text in statutes_texts.values()]
+vectorizer_statutes = TfidfVectorizer()
+tfidf_matrix_statutes = vectorizer_statutes.fit_transform(processed_statutes_texts)
+# User interaction loop
+while True:
+    user_query = input("Ask a legal-related question (type 'exit' to quit): ")
+
+    if user_query.lower() == 'exit':
+        print("Exiting the program. Goodbye!")
+        break
+
+    # Vectorize user query
+    query_vector_cases = vectorizer_cases.transform([user_query])
+    query_vector_statutes = vectorizer_statutes.transform([user_query])
+
+    # Compute cosine similarity between the query and each case
+    query_similarities_cases = cosine_similarity(query_vector_cases, tfidf_matrix_cases).flatten()
+
+    # Compute cosine similarity between the query and each statute
+    query_similarities_statutes = cosine_similarity(query_vector_statutes, tfidf_matrix_statutes).flatten()
+
+      # Retrieve the most relevant case and statute
+    top_case_idx = query_similarities_cases.argmax()
+    top_statute_idx = query_similarities_statutes.argmax()
+
+    relevant_case = cases_texts[top_case_idx]
+    relevant_statute_name = list(statutes_texts.keys())[top_statute_idx]
+    relevant_statute_content = statutes_texts[relevant_statute_name]
+    # Extract statutes
+    doc = nlp(relevant_case)
+    statutes = [ent.text for ent in doc.ents if ent.label_ == "LAW"]
+
+    # Summarize the relevant case
+    doc_sentences = list(doc.sents)
+    case_summary = "\n".join([sent.text for sent in doc_sentences])  # Include the entire case content as a summary
+
+   # Generate Statute Explanation
+    statute_explanation = f"Statute: {relevant_statute_name}\n{relevant_statute_content}"
+
+    # Generate Legal Document
+    document = f"Legal Document - User Query: {user_query}\n\n"
+    document += f"Case Summary:\n{case_summary}\n\n"
+    document += "Statute Explanation:\n"
+    document += f"{statute_explanation}\n"
+    document += "\nGuidance for the User:\n"
+    document += "To defend your friend in court, focus on presenting evidence that supports their actions were in self-defense.\n"
+    document += "Emphasize any mitigating circumstances and demonstrate their lack of intent to harm.\n"
+    document += "Consult with a qualified legal professional to build a strong defense strategy."
+
+    # Save the document to a file or display it to the user
+    with open("legal_document.txt", "w") as output_file:
+        output_file.write(document)
+
+    print("\nLegal document created. You can find the document in 'legal_document.txt'.")