open-ttg 0.0.0.4

graph builder from text

Text2Graph

🌐 Languages: 中文

---

📚 Table of Contents

• Features

• Installation

• Usage

  • Build from Concept List
  • Build from Text

• Output Format

• Testing

---

🚀 Features

• Build graphs from concept lists or raw text
• Customizable sliding window size and PMI threshold
• Returns graph data with nodes, edges, and PMI scores

---

🔧 Installation

pip install -e .

---

🧠 Usage

Build from Concept List

from text2graph.graph_builder import GraphBuilder

builder = GraphBuilder(window_size=5, pmi_threshold=0.1)
concepts = ['人工智能', '机器学习', '神经网络', '深度学习', '人工智能', '大模型', '神经网络']

nodes, edges, pmis = builder.build_graph(concepts)

Build from Text

from text2graph.graph_builder import GraphBuilder

builder = GraphBuilder(window_size=5, pmi_threshold=0.1)
text = '\n'.join(['人工智能', '机器学习', '神经网络', '深度学习', '人工智能', '大模型', '神经网络'])

data = builder.build_graph_from_text(text)
nodes = data['nodes']
edges = data['edges']
pmis = data['pmis']

---

📦 Output Format

• nodes: List of unique concepts
• edges: List of tuples (concept1, concept2)
• pmis: List of PMI scores corresponding to each edge

---

✅ Testing

python -m unittest

---

For Developers

setup docker env

docker build -t text2graph -f docker/Dockerfile .
docker run -it --rm --name text2graph -v .\:/code text2graph bash