lab9

MaastrichtU-IDS · Mar 18, 2024 · 2333139 · 2333139
1 parent c517d4e
commit 2333139
Showing 1 changed file with 82 additions and 9 deletions.
diff --git a/2024/lab9/Lab9 - RAG over KG.ipynb b/2024/lab9/Lab9 - RAG over KG.ipynb
@@ -154,7 +154,22 @@
     "\n",
     "SELECT ?uri ?predicate ?label ?type\n",
     "WHERE {\n",
-    "    TODO\n",
+    "    ?uri a ?type ;\n",
+    "        ?predicate ?label .\n",
+    "    FILTER (\n",
+    "        ?type = owl:Class ||\n",
+    "            ?type = owl:DatatypeProperty ||\n",
+    "            ?type = owl:ObjectProperty\n",
+    "    )\n",
+    "    FILTER (\n",
+    "        ?predicate = rdfs:label ||\n",
+    "            ?predicate = skos:prefLabel ||\n",
+    "            ?predicate = skos:altLabel ||\n",
+    "            ?predicate = skos:definition ||\n",
+    "            ?predicate = rdfs:comment ||\n",
+    "            ?predicate = dcterms:description ||\n",
+    "            ?predicate = dc:title\n",
+    "    )\n",
     "}\"\"\"\n",
     "\n",
     "class OntologyLoader(BaseLoader):\n",
@@ -200,9 +215,31 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 3,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\u001b[32m2024-03-17 11:02:48.072\u001b[0m | \u001b[33m\u001b[1mWARNING \u001b[0m | \u001b[36mfastembed.embedding\u001b[0m:\u001b[36m<module>\u001b[0m:\u001b[36m7\u001b[0m - \u001b[33m\u001b[1mDefaultEmbedding, FlagEmbedding, JinaEmbedding are deprecated. Use TextEmbedding instead.\u001b[0m\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "4e0eea67b54c477ea3fd7e5ecd8e7e45",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "Fetching 9 files:   0%|          | 0/9 [00:00<?, ?it/s]"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
    "source": [
     "flag_embeddings = FastEmbedEmbeddings(model_name=\"BAAI/bge-small-en-v1.5\", max_length=512)\n",
     "loader = OntologyLoader(\"https://semanticscience.org/ontology/sio.owl\", format=\"xml\")\n",
@@ -251,7 +288,10 @@
     ")\n",
     "\n",
     "# Prompt to reformulate the question using the chat history\n",
-    "reform_template = \"\"\"TODO\n",
+    "reform_template = \"\"\"Given the following chat history and a follow up question,\n",
+    "rephrase the follow up question to be a standalone straightforward question, in its original language.\n",
+    "Do not answer the question! Just rephrase reusing informations from the chat history.\n",
+    "Make it short and straight to the point.\n",
     "\n",
     "Chat History:\n",
     "{chat_history}\n",
@@ -260,7 +300,8 @@
     "REFORM_QUESTION_PROMPT = PromptTemplate.from_template(reform_template)\n",
     "\n",
     "# Prompt to ask to answer the reformulated question\n",
-    "answer_template = \"\"\"TODO\n",
+    "answer_template = \"\"\"Briefly answer the question based only on the following context,\n",
+    "do not use any information outside this context:\n",
     "{context}\n",
     "\n",
     "Question: {question}\n",
@@ -351,9 +392,25 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 10,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "💭 Reformulated question:  Can you give me the definition of a protein?\n",
+      "📚 Documents retrieved:\n",
+      "· protein (http://semanticscience.org/resource/SIO_010043)\n",
+      "· A protein is an organic polymer that is composed of one or more linear polymers of amino acids. (http://semanticscience.org/resource/SIO_010043)\n",
+      "· A protein complex is a molecular complex composed of at least two polypeptide chains. (http://semanticscience.org/resource/SIO_010497)\n",
+      "· A polypeptide is an organic polymer composed of amino acid residues, typically of less than 50 amino acids in length. (http://semanticscience.org/resource/SIO_010007)\n",
+      "· amino acid (http://semanticscience.org/resource/SIO_001224)\n",
+      "\n",
+      "Answer: A protein is an organic polymer composed of one or more linear polymers of amino acids."
+     ]
+    }
+   ],
    "source": [
     "# set_debug(True)   # Uncomment to enable detailed LangChain debugging\n",
     "output = stream_chain(final_chain, memory, {\n",
@@ -363,9 +420,25 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 11,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "💭 Reformulated question:  What is the URI for the concept of a protein?\n",
+      "📚 Documents retrieved:\n",
+      "· protein (http://semanticscience.org/resource/SIO_010043)\n",
+      "· A protein complex is a molecular complex composed of at least two polypeptide chains. (http://semanticscience.org/resource/SIO_010497)\n",
+      "· protein complex (http://semanticscience.org/resource/SIO_010497)\n",
+      "· A protein-protein association is an association between two proteins. (http://semanticscience.org/resource/SIO_001438)\n",
+      "· protein-protein association (http://semanticscience.org/resource/SIO_001438)\n",
+      "\n",
+      "http://semanticscience.org/resource/SIO_010043"
+     ]
+    }
+   ],
    "source": [
     "output = stream_chain(final_chain, memory, {\n",
     "    \"question\": \"What is the URI for this concept?\"\n",