From 795acc7c4096132f80e361d59a99f67fe8670f7c Mon Sep 17 00:00:00 2001
From: "Muhammad.haekal Rahmatullah" <haekalrahmatullah36@gmail.com>
Date: Sun, 1 Nov 2026 23:06:04 -0500
Subject: [PATCH] Upload files to "/"

---
 gemini-code-1793591350535 Haekal.py | 149 ++++++++++++++++++++++++++++
 1 file changed, 149 insertions(+)
 create mode 100644 gemini-code-1793591350535 Haekal.py

diff --git a/gemini-code-1793591350535 Haekal.py b/gemini-code-1793591350535 Haekal.py
new file mode 100644
index 0000000..fdc3ff8
--- /dev/null
+++ b/gemini-code-1793591350535 Haekal.py	
@@ -0,0 +1,149 @@
+import math
+import re
+from collections import Counter
+
+
+class UKWMSAI_RAG:
+  """Sistem RAG Sederhana berbasis Pure Python untuk topik Kecerdasan Buatan
+
+  di Teknik Informatika Universitas Katolik Widya Mandala Surabaya (UKWMS).
+  """
+
+  def __init__(self, knowledge_base):
+    print("[INFO] Menginisialisasi Basis Pengetahuan UKWMS - AI...")
+    self.documents = knowledge_base
+    self.vocab = self._build_vocab(self.documents)
+    # Pra-penghitungan vektor TF-IDF untuk dokumen lokal
+    self.doc_vectors = [
+        self._compute_tfidf(doc, self.vocab, self.documents)
+        for doc in self.documents
+    ]
+
+  def _tokenize(self, text):
+    """Membersihkan teks dan memecahnya menjadi kata-kata (token)."""
+    return re.findall(r"\b\w+\b", text.lower())
+
+  def _build_vocab(self, docs):
+    """Membangun kamus kata unik dari seluruh dokumen."""
+    vocab = set()
+    for doc in docs:
+      vocab.update(self._tokenize(doc))
+    return sorted(list(vocab))
+
+  def _compute_tfidf(self, text, vocab, all_docs):
+    """Menghitung bobot TF-IDF secara manual tanpa library eksternal."""
+    tokens = self._tokenize(text)
+    tf = Counter(tokens)
+    total_tokens = len(tokens) if tokens else 1
+
+    vector = {}
+    for word in vocab:
+      # Term Frequency (TF)
+      term_freq = tf[word] / total_tokens
+
+      # Inverse Document Frequency (IDF)
+      doc_count = sum(1 for d in all_docs if word in self._tokenize(d))
+      idf = (
+          math.log(len(all_docs) / (doc_count + 1)) + 1 if doc_count > 0 else 0
+      )
+
+      vector[word] = term_freq * idf
+    return vector
+
+  def _cosine_similarity(self, vec1, vec2):
+    """Menghitung kemiripan kosinus antara dua vektor TF-IDF."""
+    intersection = set(vec1.keys()).intersection(set(vec2.keys()))
+    numerator = sum(vec1[x] * vec2[x] for x in intersection)
+
+    sum1 = sum(val ** 2 for val in vec1.values())
+    sum2 = sum(val ** 2 for val in vec2.values())
+    denominator = math.sqrt(sum1) * math.sqrt(sum2)
+
+    if denominator == 0:
+      return 0.0
+    return numerator / denominator
+
+  def retrieve(self, query, top_k=1):
+    """Mencari dokumen paling relevan menggunakan Cosine Similarity."""
+    query_vector = self._compute_tfidf(query, self.vocab, self.documents)
+    scores = []
+
+    for i, doc_vec in enumerate(self.doc_vectors):
+      score = self._cosine_similarity(query_vector, doc_vec)
+      scores.append((score, self.documents[i]))
+
+    # Urutkan berdasarkan skor tertinggi
+    scores.sort(key=lambda x: x[0], reverse=True)
+    return [doc for score, doc in scores[:top_k]]
+
+  def generate_answer(self, query):
+    """Tahap Generation: Menggabungkan hasil retrieval dengan format jawaban RAG."""
+    retrieved_docs = self.retrieve(query, top_k=1)
+    context = (
+        retrieved_docs[0]
+        if retrieved_docs
+        else "Maaf, informasi tidak ditemukan."
+    )
+
+    # Template Generator Extractive RAG
+    answer = (
+        f"==================================================\n"
+        f"HASIL GENERASI RAG - INFORMATIKA UKWMS (KECERDASAN BUATAN)\n"
+        f"==================================================\n"
+        f"Pertanyaan : {query}\n\n"
+        f"Konteks Relevan (Retrieved):\n-> {context}\n\n"
+        f"Jawaban Sistem:\n"
+        f"Berdasarkan kurikulum dan data akademik Prodi Informatika UKWMS, "
+        f"fokus pada bidang Kecerdasan Buatan mencakup informasi berikut:\n"
+        f'"{context}"\n'
+        f"=================================================="
+    )
+    return answer, retrieved_docs
+
+
+# ==========================================
+# EKSEKUSI PROGRAM
+# ==========================================
+if __name__ == "__main__":
+  # 1. Dataset / Basis Pengetahuan Lokal UKWMS - Kecerdasan Buatan
+  ukwms_ai_kb = [
+      (
+          "Program Studi Informatika Fakultas Teknik UKWMS menyediakan"
+          " peminatan khusus Data Science and Artificial Intelligence (AI)"
+          " untuk membekali mahasiswa dengan kemampuan analisis data dan"
+          " pembelajaran mesin."
+      ),
+      (
+          "Mata kuliah inti dalam bidang Kecerdasan Buatan di Informatika UKWMS"
+          " meliputi Machine Learning, Deep Learning, Natural Language"
+          " Processing (NLP), Computer Vision, dan Data Mining."
+      ),
+      (
+          "Laboratorium Komputasi dan Sistem Cerdas di Teknik Informatika"
+          " UKWMS Kampus Kalijudan Surabaya sering digunakan mahasiswa untuk"
+          " mengembangkan proyek akhir berbasis kecerdasan buatan dan IoT."
+      ),
+      (
+          "Penelitian dosen dan mahasiswa Informatika UKWMS di bidang AI"
+          " banyak berfokus pada pengenalan pola citra medis, sistem"
+          " rekomendasi, dan analisis sentimen bahasa Indonesia."
+      ),
+      (
+          "Lulusan peminatan Kecerdasan Buatan dari Informatika UKWMS bergelar"
+          " Sarjana Komputer (S.Kom.) dengan profisiensi tinggi sebagai AI"
+          " Engineer, Data Scientist, dan Machine Learning Practitioner."
+      ),
+  ]
+
+  # 2. Inisialisasi Sistem RAG
+  rag = UKWMSAI_RAG(ukwms_ai_kb)
+
+  # 3. Contoh Pengujian Pertanyaan Tugas
+  sample_query = (
+      "Apa saja mata kuliah kecerdasan buatan yang dipelajari di informatika"
+      " ukwms?"
+  )
+
+  # 4. Menjalankan Pipeline
+  final_answer, sources = rag.generate_answer(sample_query)
+  print("\n" + final_answer)
\ No newline at end of file