├── .python-version ├── knowledge_graph_rag ├── utils │ ├── __init__.py │ ├── llm.py │ ├── prompts.py │ └── text_preprocessing.py ├── documents_cluster.py ├── vectordb.py ├── __init__.py ├── document.py ├── knowledge_graph.py └── documents_graph.py ├── assets ├── documents_graph.png └── knowledge_graph.png ├── .gitignore ├── setup.py ├── LICENSE ├── README.MD └── examples └── documents_graph_usage.ipynb /.python-version: -------------------------------------------------------------------------------- 1 | 3.12.2 2 | -------------------------------------------------------------------------------- /knowledge_graph_rag/utils/__init__.py: -------------------------------------------------------------------------------- 1 | -------------------------------------------------------------------------------- /assets/documents_graph.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sarthakrastogi/graph-rag/HEAD/assets/documents_graph.png -------------------------------------------------------------------------------- /assets/knowledge_graph.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sarthakrastogi/graph-rag/HEAD/assets/knowledge_graph.png -------------------------------------------------------------------------------- /knowledge_graph_rag/documents_cluster.py: -------------------------------------------------------------------------------- 1 | class DocumentsCluster: 2 | def __init__(self) -> None: 3 | pass 4 | -------------------------------------------------------------------------------- /knowledge_graph_rag/vectordb.py: -------------------------------------------------------------------------------- 1 | class VectorDBCollection: 2 | def __init__(self, vendor="chromadb") -> None: 3 | self.vendor = vendor 4 | -------------------------------------------------------------------------------- /knowledge_graph_rag/__init__.py: -------------------------------------------------------------------------------- 1 | from .vectordb import VectorDBCollection 2 | from .documents_cluster import DocumentsCluster 3 | from .documents_graph import DocumentsGraph 4 | -------------------------------------------------------------------------------- /knowledge_graph_rag/utils/llm.py: -------------------------------------------------------------------------------- 1 | from litellm import completion 2 | 3 | def llm_call(messages, model="gpt-3.5-turbo"): 4 | response = completion(model="gpt-3.5-turbo", messages=messages) 5 | return response.choices[0].message.content -------------------------------------------------------------------------------- /knowledge_graph_rag/document.py: -------------------------------------------------------------------------------- 1 | class Document: 2 | def __init__(self, content, embedding=[], title="", source="") -> None: 3 | self.content = content 4 | self.embedding = embedding 5 | self.title = title 6 | self.source = source 7 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | .DS_Store 2 | .env 3 | utils/__pycache__ 4 | graph-rag/__pycache__ 5 | graph-rag/utils/__pycache__ 6 | examples/__pycache__ 7 | resources/__pycache__ 8 | __pycache__ 9 | .gitmodules 10 | usage.ipynb 11 | build/ 12 | dist/ 13 | knowledge_graph_rag.egg-info/ 14 | cvd_vectors/ 15 | med_graph.pickle -------------------------------------------------------------------------------- /setup.py: -------------------------------------------------------------------------------- 1 | from setuptools import setup, find_packages 2 | 3 | with open("README.MD", "r") as f: 4 | readme_content = f.read() 5 | 6 | setup( 7 | name="knowledge_graph_rag", 8 | version="0.1.0", 9 | packages=find_packages(), 10 | long_description=readme_content, 11 | long_description_content_type="text/markdown", 12 | install_requires=[ 13 | "numpy==1.24.0", 14 | "networkx==3.2.1", 15 | "nltk==3.8.1", 16 | "litellm==1.34.0", 17 | ], 18 | ) 19 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | Copyright (c) 2024 Sarthak Rastogi 2 | 3 | Permission is hereby granted, free of charge, to any person obtaining a copy 4 | of this software and associated documentation files (the "Software"), to deal 5 | in the Software without restriction, including without limitation the rights 6 | to use, copy, modify, merge, publish, distribute, sublicense, and/or sell 7 | copies of the Software, and to permit persons to whom the Software is 8 | furnished to do so, subject to the following conditions: 9 | 10 | The above copyright notice and this permission notice shall be included in all 11 | copies or substantial portions of the Software. 12 | 13 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 14 | IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 15 | FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE 16 | AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 17 | LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, 18 | OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE 19 | SOFTWARE. -------------------------------------------------------------------------------- /knowledge_graph_rag/utils/prompts.py: -------------------------------------------------------------------------------- 1 | knowledge_graph_creation_system_prompt = """ 2 | You are given a document and your task is to create a knowledge graph from it. 3 | In the knowledge graph, entities such as people, places, objects, institutions, topics, ideas, etc. are represented as nodes. 4 | Whereas the relationships and actions between them are represented as edges. 5 | 6 | You will respond with a knowledge graph in the given JSON format: 7 | 8 | [ 9 | {"entity" : "Entity_name", "connections" : [ 10 | {"entity" : "Connected_entity_1", "relationship" : "Relationship_with_connected_entity_1}, 11 | {"entity" : "Connected_entity_2", "relationship" : "Relationship_with_connected_entity_2}, 12 | ] 13 | }, 14 | {"entity" : "Entity_name", "connections" : [ 15 | {"entity" : "Connected_entity_1", "relationship" : "Relationship_with_connected_entity_1}, 16 | {"entity" : "Connected_entity_2", "relationship" : "Relationship_with_connected_entity_2}, 17 | ] 18 | }, 19 | ] 20 | 21 | You must strictly respond in the given JSON format or your response will not be parsed correctly! 22 | """ 23 | -------------------------------------------------------------------------------- /knowledge_graph_rag/utils/text_preprocessing.py: -------------------------------------------------------------------------------- 1 | import nltk 2 | from nltk.corpus import stopwords 3 | from nltk.tokenize import word_tokenize 4 | from nltk.stem import WordNetLemmatizer 5 | 6 | # Download necessary NLTK data files 7 | nltk.download("punkt") 8 | nltk.download("stopwords") 9 | nltk.download("wordnet") 10 | nltk.download("omw-1.4") 11 | 12 | 13 | def remove_stop_words_from_and_lemmatise_documents(documents): 14 | # Initialize stop words and lemmatizer 15 | stop_words = set(stopwords.words("english")) 16 | lemmatizer = WordNetLemmatizer() 17 | 18 | # Function to preprocess text 19 | def preprocess_text(sentences): 20 | preprocessed_sentences = [] 21 | 22 | for sentence in sentences: 23 | # Tokenize the sentence 24 | words = word_tokenize(sentence) 25 | 26 | # Remove stop words and lemmatize each word 27 | filtered_words = [ 28 | lemmatizer.lemmatize(word.lower()) 29 | for word in words 30 | if word.lower() not in stop_words and word.isalpha() 31 | ] 32 | 33 | # Join words back to form the sentence 34 | preprocessed_sentence = " ".join(filtered_words) 35 | preprocessed_sentences.append(preprocessed_sentence) 36 | 37 | return preprocessed_sentences 38 | 39 | # Preprocess the list of sentences 40 | preprocessed_documents = preprocess_text(documents) 41 | return preprocessed_documents 42 | -------------------------------------------------------------------------------- /README.MD: -------------------------------------------------------------------------------- 1 | # Knowledge Graph RAG 2 | ## Automatically create knowledge graphs + document networks to boost performance on RAG 3 | 4 | ### 1. Install Knowledge Graph RAG: 5 | 6 | `pip install knowledge_graph_rag` 7 | 8 | ### 2. Create a Knowledge Graph or a Document Graph: 9 | 10 | ``` 11 | # Creating KG on medical documents 12 | documents = ["Cardiovascular disease ...", 13 | "Emerging therapeutic interventions ...", 14 | "The epidemiological burden ... 15 | "Cardiovascular disease also ...", 16 | "Advanced imaging techniques, ...", 17 | "Role of novel biomarkers ..." 18 | ] 19 | knowledge_graph = KnowledgeGraph(documents) 20 | knowledge_graph.create() 21 | knowledge_graph.plot() 22 | ``` 23 | ![Knowledge graph](assets/knowledge_graph.png) 24 | 25 | ``` 26 | documents_graph = DocumentsGraph(documents=documents) 27 | documents_graph.plot() 28 | ``` 29 | ![Documents graph](assets/documents_graph.png) 30 | 31 | ### 3. Search knowledge graph entities or find interconnected documents, to augment your LLM context: 32 | 33 | ``` 34 | knowledge_graph.search_document(user_query) 35 | ``` 36 | 37 | ``` 38 | >> Entity: cardiovascular disease 39 | -> antihypertensive agents (Relationship: involves treatment with) 40 | -> statins (Relationship: used to modulate dyslipidemia) 41 | -> antiplatelet therapy (Relationship: utilized to mitigate thrombosis risk) 42 | -> biomarkers (Relationship: detection and prognostication of acute coronary syndromes and heart failure) 43 | -> high-sensitivity troponins (Relationship: detection of acute coronary syndromes and heart failure) 44 | -> natriuretic peptides (Relationship: prognostication of acute coronary syndromes and heart failure) 45 | ``` 46 | 47 | 48 | 49 | ``` 50 | documents_containing_connected_terminology = documents_graph.find_connected_documents(vectordb_search_result) 51 | documents_containing_connected_terminology 52 | ``` 53 | 54 | ``` 55 | >> [{'document': 'emerging therapeutic intervention ...'}, 56 | {'document': 'management cardiovascular ...'}, 57 | {'document': 'role novel biomarkers ...'}] 58 | ``` 59 | 60 | 61 | ## Star History 62 | 63 | [![Star History Chart](https://api.star-history.com/svg?repos=sarthakrastogi/graph-rag&type=Date)](https://star-history.com/#sarthakrastogi/graph-rag&Date) -------------------------------------------------------------------------------- /knowledge_graph_rag/knowledge_graph.py: -------------------------------------------------------------------------------- 1 | from knowledge_graph_rag.utils.llm import llm_call 2 | import json 3 | import re 4 | from tqdm.notebook import tqdm 5 | import networkx as nx 6 | from collections import defaultdict, deque 7 | import matplotlib.pyplot as plt 8 | 9 | from .utils.prompts import knowledge_graph_creation_system_prompt 10 | 11 | 12 | class KnowledgeGraph: 13 | def __init__(self, documents) -> None: 14 | self.documents = documents 15 | 16 | def remove_trailing_commas(self, json_string): 17 | # Remove trailing commas from JSON arrays and objects 18 | json_string = re.sub(r",\s*([\]}])", r"\1", json_string) 19 | return json_string 20 | 21 | def create_knowledge_representations(self, documents): 22 | knowledge_representations_of_individual_documents = [] 23 | for document in tqdm(documents): 24 | messages = [ 25 | {"role": "system", "content": knowledge_graph_creation_system_prompt}, 26 | {"role": "user", "content": document}, 27 | ] 28 | 29 | response = llm_call(messages=messages) 30 | response = response.lower() 31 | response = self.remove_trailing_commas(response) 32 | knowledge_representations_of_individual_documents.append( 33 | json.loads(response) 34 | ) 35 | 36 | return knowledge_representations_of_individual_documents 37 | 38 | def create_knowledge_graph_from_representations(self, representations): 39 | G = nx.DiGraph() 40 | 41 | def add_edge(source, target, relationship): 42 | if G.has_edge(source, target): 43 | G[source][target]["relationship"] += f", {relationship}" 44 | G[source][target]["weight"] = G[source][target].get("weight", 1) + 1 45 | else: 46 | G.add_edge(source, target, relationship=relationship, weight=1) 47 | 48 | for rep in representations: 49 | for item in rep: 50 | source = item["entity"] 51 | if "connections" in item: 52 | for conn in item["connections"]: 53 | target = conn["entity"] 54 | relationship = conn["relationship"] 55 | add_edge(source, target, relationship) 56 | 57 | return G 58 | 59 | def create(self): 60 | self.knowledge_representations = self.create_knowledge_representations( 61 | self.documents 62 | ) 63 | self.G = self.create_knowledge_graph_from_representations( 64 | self.knowledge_representations 65 | ) 66 | 67 | def plot(self): 68 | pos = nx.spring_layout(self.G) 69 | plt.figure(figsize=(12, 8)) 70 | 71 | # Draw nodes with labels 72 | nx.draw_networkx_nodes( 73 | self.G, pos, node_size=5000, node_color="skyblue", alpha=0.7 74 | ) 75 | node_labels = { 76 | node: node[:20] + "..." if len(node) > 20 else node 77 | for node in self.G.nodes() 78 | } 79 | nx.draw_networkx_labels( 80 | self.G, pos, labels=node_labels, font_size=10, font_family="sans-serif" 81 | ) 82 | 83 | # Draw edges with weights 84 | edges = self.G.edges(data=True) 85 | for u, v, d in edges: 86 | weight = d.get("weight", 1) # Default to 1 if weight is not present 87 | nx.draw_networkx_edges( 88 | self.G, pos, edgelist=[(u, v)], width=weight, alpha=0.5 89 | ) 90 | 91 | # Add edge labels (relationship and weight) 92 | edge_label = ( 93 | f"{d['relationship'][:20]}...\n(w:{weight:.2f})" 94 | if len(d["relationship"]) > 20 95 | else f"{d['relationship']}\n(w:{weight:.2f})" 96 | ) 97 | x = (pos[u][0] + pos[v][0]) / 2 98 | y = (pos[u][1] + pos[v][1]) / 2 99 | plt.text( 100 | x, 101 | y, 102 | edge_label, 103 | fontsize=8, 104 | ha="center", 105 | va="center", 106 | bbox=dict(facecolor="white", edgecolor="none", alpha=0.7), 107 | ) 108 | 109 | plt.axis("off") 110 | plt.tight_layout() 111 | plt.show() 112 | 113 | def search_document(self, input_document, max_depth=3): 114 | knowledge_representations_of_input_document = ( 115 | self.create_knowledge_representations(documents=[input_document]) 116 | ) 117 | result = [] 118 | for rep in knowledge_representations_of_input_document: 119 | for item in rep: 120 | source_entity = item["entity"] 121 | if source_entity in self.G: 122 | result.append(f"\nEntity: {source_entity}") 123 | result.extend(self.bfs_traversal(source_entity, max_depth)) 124 | return "\n".join(result) 125 | 126 | def bfs_traversal(self, start_node, max_depth): 127 | visited = set() 128 | queue = deque([(start_node, 0)]) 129 | result = [] 130 | while queue: 131 | node, depth = queue.popleft() 132 | if depth > max_depth: 133 | break 134 | if node not in visited: 135 | visited.add(node) 136 | for neighbor in self.G.neighbors(node): 137 | if neighbor not in visited: 138 | relationship = self.G[node][neighbor]["relationship"] 139 | result.append(f" -> {neighbor} (Relationship: {relationship})") 140 | queue.append((neighbor, depth + 1)) 141 | return result 142 | -------------------------------------------------------------------------------- /knowledge_graph_rag/documents_graph.py: -------------------------------------------------------------------------------- 1 | import numpy as np 2 | import networkx as nx 3 | import matplotlib.pyplot as plt 4 | from sklearn.feature_extraction.text import TfidfVectorizer 5 | from sklearn.metrics.pairwise import cosine_similarity 6 | import pickle 7 | 8 | from knowledge_graph_rag.utils.text_preprocessing import ( 9 | remove_stop_words_from_and_lemmatise_documents, 10 | ) 11 | 12 | 13 | class DocumentsGraph: 14 | def __init__(self, documents) -> None: 15 | self.documents = documents 16 | self.preprocessed_documents = remove_stop_words_from_and_lemmatise_documents( 17 | documents=documents 18 | ) 19 | self.G = self.create_graph_from_documents() 20 | 21 | def create_graph_from_documents(self): 22 | # Compute TF-IDF 23 | vectorizer = TfidfVectorizer() 24 | tfidf_matrix = vectorizer.fit_transform(self.preprocessed_documents) 25 | 26 | # Compute cosine similarity matrix 27 | cosine_sim = cosine_similarity(tfidf_matrix) 28 | 29 | # Create the graph 30 | G = nx.Graph() 31 | 32 | # Add nodes 33 | for i, doc in enumerate(self.preprocessed_documents): 34 | G.add_node(i, label=doc) 35 | 36 | # Add edges with weights (cosine similarity) 37 | for i in range(len(self.preprocessed_documents)): 38 | for j in range(i + 1, len(self.preprocessed_documents)): 39 | weight = cosine_sim[i, j] 40 | if weight > 0: # Add edge only if there's a similarity 41 | G.add_edge(i, j, weight=weight) 42 | 43 | return G 44 | 45 | def plot(self): 46 | # Draw the graph with labels and edge weights 47 | pos = nx.spring_layout(self.G) 48 | 49 | plt.figure(figsize=(12, 8)) 50 | 51 | # Draw nodes with labels 52 | node_labels = nx.get_node_attributes(self.G, "label") 53 | node_labels = { 54 | node_number: node_label[:20] + "..." 55 | for node_number, node_label in node_labels.items() 56 | } 57 | nx.draw_networkx_nodes( 58 | self.G, pos, node_size=5000, node_color="skyblue", alpha=0.7 59 | ) 60 | nx.draw_networkx_labels( 61 | self.G, pos, labels=node_labels, font_size=10, font_family="sans-serif" 62 | ) 63 | 64 | # Draw edges with weights 65 | edges = self.G.edges(data=True) 66 | for u, v, d in edges: 67 | weight = d["weight"] 68 | nx.draw_networkx_edges( 69 | self.G, pos, edgelist=[(u, v)], width=weight * 10, alpha=0.5 70 | ) 71 | edge_label = f"{weight:.4f}" 72 | mid_edge = (pos[u] + pos[v]) / 2 73 | plt.text( 74 | mid_edge[0], 75 | mid_edge[1], 76 | edge_label, 77 | fontsize=9, 78 | ha="center", 79 | va="center", 80 | ) 81 | 82 | plt.axis("off") 83 | plt.show() 84 | 85 | def find_connected_documents(self, input_sentence, N=3): 86 | # Find the node corresponding to the given sentence 87 | input_sentence = remove_stop_words_from_and_lemmatise_documents( 88 | documents=[input_sentence] 89 | )[0] 90 | node_index = None 91 | for node, data in self.G.nodes(data=True): 92 | if data["label"] == input_sentence: 93 | node_index = node 94 | break 95 | 96 | if node_index is None: 97 | raise ValueError("The provided sentence is not in the graph.") 98 | 99 | # Get the neighbors and their edge weights 100 | neighbors = [ 101 | (neighbor, self.G[node_index][neighbor]["weight"]) 102 | for neighbor in self.G.neighbors(node_index) 103 | ] 104 | 105 | # Sort neighbors by edge weight in descending order 106 | neighbors = sorted(neighbors, key=lambda x: x[1], reverse=True) 107 | 108 | # Return the top N neighbors with their full text and weights 109 | top_neighbors = [ 110 | {"document": self.G.nodes[neighbor]["label"]} # , "similarity": weight} 111 | for neighbor, weight in neighbors[:N] 112 | ] 113 | return top_neighbors 114 | 115 | def find_k_closest_sentences(self, input_sentence, N=5): 116 | input_sentence = remove_stop_words_from_and_lemmatise_documents( 117 | documents=[input_sentence] 118 | )[0] 119 | 120 | # Append the input_sentence to the list of documents 121 | all_docs = self.preprocessed_documents + [input_sentence] 122 | 123 | # Compute TF-IDF for all documents including the input sentence 124 | vectorizer = TfidfVectorizer() 125 | tfidf_matrix = vectorizer.fit_transform(all_docs) 126 | 127 | # Compute cosine similarity between all pairs of documents 128 | cosine_sim = cosine_similarity(tfidf_matrix) 129 | similarity_scores = cosine_sim[-1, :-1] # Exclude the similarity with itself 130 | 131 | # Get the indices of the top N similar documents 132 | closest_indices = np.argsort(similarity_scores)[-N:][::-1] 133 | 134 | # Return the closest N sentences and their similarity scores 135 | closest_sentences = [ 136 | (self.preprocessed_documents[idx], similarity_scores[idx]) 137 | for idx in closest_indices 138 | if similarity_scores[idx] > 0 139 | ] 140 | return closest_sentences 141 | 142 | def save(self, graph_name): 143 | # save graph object to file 144 | pickle.dump(self.G, open(f"{graph_name}.pickle", "wb")) 145 | 146 | def load_from_file(self, graph_name): 147 | # load graph object from file 148 | self.G = pickle.load(open(f"{graph_name}.pickle", "rb")) 149 | -------------------------------------------------------------------------------- /examples/documents_graph_usage.ipynb: -------------------------------------------------------------------------------- 1 | { 2 | "cells": [ 3 | { 4 | "cell_type": "markdown", 5 | "metadata": {}, 6 | "source": [ 7 | "\n", 8 | " \"Open\n", 9 | "" 10 | ] 11 | }, 12 | { 13 | "cell_type": "code", 14 | "execution_count": 1, 15 | "metadata": {}, 16 | "outputs": [ 17 | { 18 | "data": { 19 | "text/plain": [ 20 | "'\\n%%capture\\n!pip install knowledge_graph_rag\\n!pip install numpy==1.24.0\\n!pip install chromadb\\n'" 21 | ] 22 | }, 23 | "execution_count": 1, 24 | "metadata": {}, 25 | "output_type": "execute_result" 26 | } 27 | ], 28 | "source": [ 29 | "%%capture\n", 30 | "!pip install knowledge_graph_rag\n", 31 | "!pip install numpy==1.24.0\n", 32 | "!pip install chromadb" 33 | ] 34 | }, 35 | { 36 | "cell_type": "code", 37 | "execution_count": 2, 38 | "metadata": {}, 39 | "outputs": [], 40 | "source": [ 41 | "%%capture\n", 42 | "from knowledge_graph_rag.document import Document\n", 43 | "from knowledge_graph_rag.documents_graph import DocumentsGraph" 44 | ] 45 | }, 46 | { 47 | "cell_type": "code", 48 | "execution_count": 3, 49 | "metadata": {}, 50 | "outputs": [], 51 | "source": [ 52 | "\n", 53 | "documents = [\"Cardiovascular disease (CVD) encompasses a spectrum of disorders involving the heart and vasculature, prominently including atherosclerosis, characterized by endothelial dysfunction and the accumulation of lipid-laden plaques. These pathophysiological processes often precipitate myocardial infarction and cerebrovascular accidents, arising from the rupture of vulnerable plaques and subsequent thrombogenesis.\",\n", 54 | " \"Management of cardiovascular disease necessitates a multifaceted approach involving antihypertensive agents, statins to modulate dyslipidemia, and antiplatelet therapy to mitigate thrombosis risk.\",\n", 55 | " \"Emerging therapeutic interventions targeting molecular pathways, including PCSK9 inhibitors and SGLT2 inhibitors, show promise in reducing cardiovascular morbidity and mortality.\",\n", 56 | " \"The epidemiological burden of cardiovascular disease underscores the imperative for ongoing research into genetic predispositions and the optimization of primary and secondary prevention strategies.\"\n", 57 | " \"Cardiovascular disease also significantly intersects with metabolic syndrome, wherein insulin resistance and visceral adiposity contribute to endothelial dysfunction and systemic inflammation, further accelerating atherogenic processes.\",\n", 58 | " \"Advanced imaging techniques, such as coronary artery calcium scoring and carotid intima-media thickness measurement, enhance the stratification of cardiovascular risk, enabling more tailored therapeutic interventions.\",\n", 59 | " \"Role of novel biomarkers, including high-sensitivity troponins and natriuretic peptides, is pivotal in the early detection and prognostication of acute coronary syndromes and heart failure within the broader spectrum of cardiovascular disease.\"\n", 60 | "]" 61 | ] 62 | }, 63 | { 64 | "cell_type": "markdown", 65 | "metadata": {}, 66 | "source": [ 67 | "Graph RAG can perform much better than std RAG. Here’s when and how:\n", 68 | "\n", 69 | "When you want your LLM to understand the interconnection between your documents before arriving to its answer, Graph RAG is necessary.\n", 70 | "\n", 71 | "RAG returns search results based on semantic similarity. It fails to consider that, if doc A is selected as highly relevant, the docs containing data closely linked to doc A must be included in the context to give a full picture.\n", 72 | "\n", 73 | "This is where we need Graph RAG.\n", 74 | "\n", 75 | "Search results from a graph are more likely to give you a comprehensive view of the entity being searched and the info connected to it.\n", 76 | "\n", 77 | "Information on entities like people, institutions, etc. is often highly interconnected, and this might be the case for your data too.\n" 78 | ] 79 | }, 80 | { 81 | "cell_type": "markdown", 82 | "metadata": {}, 83 | "source": [ 84 | "### 1. Create a VectorDB" 85 | ] 86 | }, 87 | { 88 | "cell_type": "code", 89 | "execution_count": 4, 90 | "metadata": {}, 91 | "outputs": [], 92 | "source": [ 93 | "def get_embedding_batch(input_array):\n", 94 | " from openai import OpenAI\n", 95 | " client = OpenAI(api_key=\"YOUR_OPENAI_API_KEY\")\n", 96 | " response = client.embeddings.create(\n", 97 | " input=input_array,\n", 98 | " model=\"text-embedding-3-small\"\n", 99 | " )\n", 100 | " return [data.embedding for data in response.data]" 101 | ] 102 | }, 103 | { 104 | "cell_type": "code", 105 | "execution_count": 5, 106 | "metadata": {}, 107 | "outputs": [], 108 | "source": [ 109 | "embeddings = get_embedding_batch(documents)\n", 110 | "vectors_collection = [{document : embedding} for document, embedding in zip(documents, embeddings)]" 111 | ] 112 | }, 113 | { 114 | "cell_type": "code", 115 | "execution_count": 6, 116 | "metadata": {}, 117 | "outputs": [], 118 | "source": [ 119 | "import chromadb\n", 120 | "vectordb_name = \"cvd_vectors\"\n", 121 | "client = chromadb.PersistentClient(path=vectordb_name)\n", 122 | "collection = client.create_collection(vectordb_name)\n", 123 | "\n", 124 | "collection.add(\n", 125 | " embeddings=embeddings,\n", 126 | " documents=documents,\n", 127 | " metadatas=[{\"source\" : \"\"} for i in range(len(documents))],\n", 128 | " ids=list(map(str, range(len(documents))))\n", 129 | ")" 130 | ] 131 | }, 132 | { 133 | "cell_type": "markdown", 134 | "metadata": {}, 135 | "source": [ 136 | "### 2. Create a Documents Graph" 137 | ] 138 | }, 139 | { 140 | "cell_type": "code", 141 | "execution_count": 7, 142 | "metadata": {}, 143 | "outputs": [ 144 | { 145 | "data": { 146 | "image/png": "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", 147 | "text/plain": [ 148 | "
" 149 | ] 150 | }, 151 | "metadata": {}, 152 | "output_type": "display_data" 153 | } 154 | ], 155 | "source": [ 156 | "documents_graph = DocumentsGraph(documents=documents)\n", 157 | "documents_graph.plot()" 158 | ] 159 | }, 160 | { 161 | "cell_type": "code", 162 | "execution_count": 8, 163 | "metadata": {}, 164 | "outputs": [], 165 | "source": [ 166 | "documents_graph.save(\"med_graph\")" 167 | ] 168 | }, 169 | { 170 | "cell_type": "code", 171 | "execution_count": 9, 172 | "metadata": {}, 173 | "outputs": [], 174 | "source": [ 175 | "user_query = \"How do advanced imaging techniques enhance cardiovascular risk stratification?\"\n", 176 | "query_embeddings = get_embedding_batch(user_query)" 177 | ] 178 | }, 179 | { 180 | "cell_type": "markdown", 181 | "metadata": {}, 182 | "source": [ 183 | "### 3. Search vectorDB" 184 | ] 185 | }, 186 | { 187 | "cell_type": "code", 188 | "execution_count": 10, 189 | "metadata": {}, 190 | "outputs": [ 191 | { 192 | "name": "stdout", 193 | "output_type": "stream", 194 | "text": [ 195 | "Advanced imaging techniques, such as coronary artery calcium scoring and carotid intima-media thickness measurement, enhance the stratification of cardiovascular risk, enabling more tailored therapeutic interventions.\n" 196 | ] 197 | } 198 | ], 199 | "source": [ 200 | "vectordb_search_result = collection.query(query_embeddings=query_embeddings, n_results=1)['documents'][0][0]\n", 201 | "print(vectordb_search_result)" 202 | ] 203 | }, 204 | { 205 | "cell_type": "markdown", 206 | "metadata": {}, 207 | "source": [ 208 | "### 4. Search Documents Graph\n", 209 | "\n", 210 | "To find interconnected documents containing terminology / n-grams used in search result.\n", 211 | "\n", 212 | "Search results from a graph can give you a comprehensive view of the entity being searched and the info connected to it.\n" 213 | ] 214 | }, 215 | { 216 | "cell_type": "code", 217 | "execution_count": 11, 218 | "metadata": {}, 219 | "outputs": [ 220 | { 221 | "data": { 222 | "text/plain": [ 223 | "[{'document': 'emerging therapeutic intervention targeting molecular pathway including inhibitor inhibitor show promise reducing cardiovascular morbidity mortality'},\n", 224 | " {'document': 'management cardiovascular disease necessitates multifaceted approach involving antihypertensive agent statin modulate dyslipidemia antiplatelet therapy mitigate thrombosis risk'},\n", 225 | " {'document': 'role novel biomarkers including troponins natriuretic peptide pivotal early detection prognostication acute coronary syndrome heart failure within broader spectrum cardiovascular disease'}]" 226 | ] 227 | }, 228 | "execution_count": 11, 229 | "metadata": {}, 230 | "output_type": "execute_result" 231 | } 232 | ], 233 | "source": [ 234 | "documents_containing_connected_terminology = documents_graph.find_connected_documents(vectordb_search_result)\n", 235 | "documents_containing_connected_terminology" 236 | ] 237 | }, 238 | { 239 | "cell_type": "markdown", 240 | "metadata": {}, 241 | "source": [ 242 | "### 5. Augment interconnected documents into context" 243 | ] 244 | } 245 | ], 246 | "metadata": { 247 | "kernelspec": { 248 | "display_name": "base", 249 | "language": "python", 250 | "name": "python3" 251 | }, 252 | "language_info": { 253 | "codemirror_mode": { 254 | "name": "ipython", 255 | "version": 3 256 | }, 257 | "file_extension": ".py", 258 | "mimetype": "text/x-python", 259 | "name": "python", 260 | "nbconvert_exporter": "python", 261 | "pygments_lexer": "ipython3", 262 | "version": "3.12.2" 263 | } 264 | }, 265 | "nbformat": 4, 266 | "nbformat_minor": 2 267 | } 268 | --------------------------------------------------------------------------------