Python Setup

Python setup for the professors example using Qwen Embedding Model

This section will cover the Python code to connect to HelixDB and create nodes and edges, create embeddings for the professors, search by embeddings, and filtering.

Step 1: Setting up the environment

pip

python -m venv venv
source venv/bin/activate
pip install helix-py sentence-transformers

Step 2: Imports and Setup for HelixDB

import helix
from sentence_transformers import SentenceTransformer

model = SentenceTransformer("Qwen/Qwen3-Embedding-0.6B")
db = helix.Client(local=True, port=6969, verbose=True)

Step 3: Creating the nodes that we will link professors to

Here we start creating all the nodes that we will link a professors to.

You can also make a query in queries.hx to get the ID of the research area if you have the name, but we will store it in Python for simplicity.

Sample Data

professor_data = {
  "name": "James",
  "bio": "James is an Assistant Professor focusing on basketball analytics, computer vision, and machine learning. Projects include ShotNet and DunkGPT.",
  "department": "Computer Science",
  "university": "Uni X",
  "key_research_areas": [
    {"area": "Computer Vision for Basketball"},
    {"area": "Predictive Modelling & Simulation"},
    {"area": "Sports Analytics with Large Language Models"},
    {"area": "Wearable Sensor Data Mining"},
    {"area": "Fairness & Ethics in Sports AI"}
  ]
}

Create Professor Node

james_node =db.query("create_professor", {"name": professor_data["name"], "bio": professor_data["bio"]})
// store the ID of the professor node
james_node_id = james_node[0]['professor']["id"]

Create Department Node

computer_science_node = db.query("create_department", {"name": professor_data["department"]})
// store the ID of the department node
computer_science_node_id = computer_science_node[0]['department']["id"]

Create University Node

university_node = db.query("create_university", {"name": professor_data["university"]})
// store the ID of the university node
university_node_id = university_node[0]['university']["id"]

Create Research Area Nodes

research_area_node_ids = {}
for area in professor_data["key_research_areas"]:
    research_area_node = db.query("create_research_area", {"name": area["area"]})
    pprint(research_area_node)
    research_area_node_id = research_area_node[0]['research_area']["id"]
    research_area_node_ids[area["area"]] = research_area_node_id

Step 4: Link nodes to the professor

Link Professor to Department

db.query("link_professor_to_department", {"professor_id": james_node_id, "department_id": computer_science_node_id})

Link Professor to University

db.query("link_professor_to_university", {"professor_id": james_node_id, "university_id": university_node_id})

Link Professor to Research Area

for area in professor_data["key_research_areas"]:
    db.query("link_professor_to_research_area", {"professor_id": james_node_id, "research_area_id": research_area_node_ids[area["area"]]})

Step 5: Create Embeddings for the professor

Create Research Area Embedding

for area in professor_data["key_research_areas"]:
    embedding = model.encode(area["area"]).astype(float).tolist()
    db.query("create_research_area_embedding", {"professor_id": james_node_id, "research_area": area["area"], "vector": embedding})

Step 6: Lets answer some of the questions we asked in the beginning

1. “What professors does X research area?”

Python

db.query("get_professor_by_research_area_name", {"research_area_name": "Computer Vision for Basketball"})

2. “What professors are working in X University?”

Python

QUERY get_professors_by_university_name (university_name: String) =>
    professors <- N<Professor>::WHERE(EXISTS(_::Out<HasUniversity>::WHERE(_::{name}::EQ(university_name))))
    RETURN professors

3. “What professors are working in X Department?”

Python

db.query("get_professors_by_department_name", {"department_name": "Computer Science"})

4. “What professors are working in the X University and are working in X department?”

Python

db.query("get_professors_by_university_and_department_name", {"university_name": "Uni X", "department_name": "Computer Science"})

5. “Find me professors doing X research area”

Python

query_vector = model.encode("Computer Vision for Basketball").astype(float).tolist()
db.query("search_similar_professors_by_research_area", {"query_vector": query_vector, "k": 1})

6. “I want to find professors working in computer vision, in X university”

Python

query_vector = model.encode("Computer Vision for Basketball").astype(float).tolist()
db.query("get_professor_semantically_and_by_university_name", {"query_vector": query_vector, "university_name": "Uni X", "k": 1})