A PoC Corporation Relationship Knowledge Graph System on top of Nebula Graph.

Corp-Rel is a PoC of Corpartion Relationship Knowledge Graph System. It's built on top of the Open Source Graph Database: Nebula Graph with a dataset from nebula-shareholding-example.

corp-rel-capture.mov

Quick Start

First, please setup a Nebula Graph Cluster with data loaded from nebula-shareholding-example.

Then, clone this project:

git clone https://github.com/wey-gu/nebula-corp-rel-search.git
cd nebula-corp-rel-search

Start the backend:

python3 -m pip install -r requirements.txt
cd corp-rel-backend
export NG_ENDPOINTS="192.168.123.456:9669" # This should be your Nebula Graph Cluster GraphD Endpoint
python3 app.py

Start the frontend in another terminal:

npm install -g @vue/cli
cd nebula-corp-rel-search/corp-rel-frontend
vue serve src/main.js

Start a reverse Proxy to enable Corp-Rel Backend being served with same origin of Frontend:

For example below is a Nginx config to make :8081/ go to http://localhost:8080 and :8081/api go to http://192.168.123.456:5000/api.

http {
    include       mime.types;
    default_type  application/octet-stream;

    keepalive_timeout  65;

    server {
        listen       8081;
        server_name  localhost;
        # frontend
        location / {
            proxy_pass http://localhost:8080;
        }
        # backend
        location /api {
            proxy_pass http://192.168.123.456:5000/api;
        }
    }
#...

After above reverse proxy being configured, let's verify it via cURL:

curl --header "Content-Type: application/json" \
     --request POST \
     --data '{"entity": "c_132"}' \
     http://localhost:8081/api | jq

If it's properly responded, hen we could go to http://localhost:8081 from the web browser :).

Design Log

data from Backend Side

Backend should query node's relationship path as follow:

MATCH p=(v)-[e:hold_share|:is_branch_of|:reletive_with|:role_as*1..3]-(v2) \
WHERE id(v) IN ["c_132"] RETURN p LIMIT 100

An example of the query will be like this:

([email protected]) [shareholding]> MATCH p=(v)-[e:hold_share|:is_branch_of|:reletive_with|:role_as*1..3]-(v2) \
                           -> WHERE id(v) IN ["c_132"] RETURN p LIMIT 100
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| p                                                                                                                                                                                                                                        |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 0.0}]-("c_245" :corp{name: "Thompson-King"})>                                                                                                                             |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 3.0}]-("p_1039" :person{name: "Christian Miller"})>                                                                                                                       |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 3.0}]-("p_1399" :person{name: "Sharon Gonzalez"})>                                                                                                                        |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 9.0}]-("p_1767" :person{name: "Dr. David Vance"})>                                                                                                                        |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 11.0}]-("p_1997" :person{name: "Glenn Reed"})>                                                                                                                            |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| <("c_132" :corp{name: "Chambers LLC"})<-[:[email protected] {share: 14.0}]-("p_2341" :person{name: "Jessica Baker"})>                                                                                                                         |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
...

Leveraging nebula2-python, we could have result in below data structure:

$ python3 -m pip install nebula2-python==2.5.0
$ ipython
In [1]: from nebula2.gclient.net import ConnectionPool
In [2]: from nebula2.Config import Config
In [3]: config = Config()
   ...: config.max_connection_pool_size = 10
   ...: # init connection pool
   ...: connection_pool = ConnectionPool()
   ...: # if the given servers are ok, return true, else return false
   ...: ok = connection_pool.init([('192.168.8.137', 9669)], config)
   ...: session = connection_pool.get_session('root', 'nebula')
[2021-10-13 13:44:24,242]:Get connection to ('192.168.8.137', 9669)

In [4]: resp = session.execute("use shareholding")
In [5]: query = '''
   ...: MATCH p=(v)-[e:hold_share|:is_branch_of|:reletive_with|:role_as*1..3]-(v2) \
   ...: WHERE id(v) IN ["c_132"] RETURN p LIMIT 100
   ...: '''
In [6]: resp = session.execute(query) # Note: after nebula graph 2.6.0, we could use execute_json as well

In [7]: resp.col_size()
Out[7]: 1

In [9]: resp.row_size()
Out[10]: 100

As we know the result is actually a nebula-python path type, they could be extracted as follow with .nodes() and .relationships():

In [11]: p=resp.row_values(22)[0].as_path()

In [12]: p.nodes()
Out[12]:
[("c_132" :corp{name: "Chambers LLC"}),
 ("p_4000" :person{name: "Colton Bailey"})]

In [13]: p.relationships()
Out[13]: [("p_4000")-[:role_as@0{role: "Editorial assistant"}]->("c_132")]

For relationships/edges, we could call its .edge_name(), .properties(), .start_vertex_id(), .end_vertex_id():

In [14]: rel=p.relationships()[0]

In [15]: rel
Out[15]: ("p_4000")-[:role_as@0{role: "Editorial assistant"}]->("c_132")

In [16]: rel.edge_name()
Out[16]: 'role_as'

In [17]: rel.properties()
Out[17]: {'role': "Editorial assistant"}

In [18]: rel.start_vertex_id()
Out[18]: "p_4000"

In [19]: rel.end_vertex_id()
Out[19]: "c_132"

And for nodes/vertices, we could call its .tags(), properties, get_id():

In [20]: node=p.nodes()[0]

In [21]: node.tags()
Out[21]: ['corp']

In [22]: node.properties('corp')
Out[22]: {'name': "Chambers LLC"}

In [23]: node.get_id()
Out[23]: "c_132"

Data visualization

For the frontend, we could create a view by leveraging vue-network-d3:

npm install vue-network-d3 --save
touch src/App.vue
touch src/main.js

In src/App.vue, we create a Network instance and fill in the nodeList, and linkList fetched from backend, in below example, we put fake data as:

nodes: [
        {"id": "c_132", "name": "Chambers LLC", "tag": "corp"},
        {"id": "p_4000", "name": "Colton Bailey", "tag": "person"}],
relationships: [
        {"source": "p_4000", "target": "c_132", "properties": { "role": "Editorial assistant" }, "edge": "role_as"}]

And the full example of src/App.vue will be:

<template>
  <div id="app">
    <network
      :nodeList="nodes"
      :linkList="relationships"
      :nodeSize="nodeSize"
      :linkWidth="linkWidth"
      :linkDistance="linkDistance"
      :linkTextFrontSize="linkTextFrontSize"
      :nodeTypeKey="nodeTypeKey"
      :linkTypeKey="linkTypeKey"
      :nodeTextKey="nodeTextKey"
      :linkTextKey="linkTextKey"
      :showNodeText="showNodeText"
      :showLinkText="showLinkText"
      >
    </network>
  </div>
</template>

<script>
import Network from "vue-network-d3";

export default {
  name: "app",
  components: {
    Network
  },
  data() {
    return {
      nodes: [
        {"id": "c_132", "name": "Chambers LLC", "tag": "corp"},
        {"id": "p_4000", "name": "Colton Bailey", "tag": "person"}
      ],
      relationships: [
        {"source": "p_4000", "target": "c_132", "properties": { "role": "Editorial assistant" }, "edge": "role_as"}
      ],
      nodeSize: 18,
      linkDistance: 120,
      linkWidth: 6,
      linkTextFrontSize: 20,
      nodeTypeKey: "tag",
      linkTypeKey: "edge",
      nodeTextKey: "name",
      linkTextKey: "properties",
      showNodeText: true,
      showLinkText: true
    };
  },
};
</script>

<style>
body {
  margin: 0;
}
</style>

Together with src/main.js:

import Vue from 'vue'
import App from './App.vue'

Vue.config.productionTip = false

new Vue({
  render: h => h(App),
}).$mount('#app')

Then we could run: vue serve src/main.js to have this renderred:

The data construction in Back End:

Thus we shoud know that if the backend provides list of nodes and relationships in JSON as the following, things are perfectly connected!

Nodes:

[{"id": "c_132", "name": "Chambers LLC", "tag": "corp"},
 {"id": "p_4000", "name": "Colton Bailey", "tag": "person"}]

Relationships:

[{"source": "p_4000", "target": "c_132", "properties": { "role": "Editorial assistant" }, "edge": "role_as"},
 {"source": "p_1039", "target": "c_132", "properties": { "share": "3.0" }, "edge": "hold_share"}]

We could construct it as:

def make_graph_response(resp) -> dict:
    nodes, relationships = list(), list()
    for row_index in range(resp.row_size()):
        path = resp.row_values(row_index)[0].as_path()
        _nodes = [
            {
                "id": node.get_id(), "tag": node.tags()[0],
                "name": node.properties(node.tags()[0]).get("name", "")
                }
                for node in path.nodes()
        ]
        nodes.extend(_nodes)
        _relationships = [
            {
                "source": rel.start_vertex_id(),
                "target": rel.end_vertex_id(),
                "properties": rel.properties(),
                "edge": rel.edge_name()
                }
                for rel in path.relationships()
        ]
        relationships.extend(_relationships)
    return {"nodes": nodes, "relationships": relationships}

The Flask App

Then Let's create a Flask App to consume the HTTP API request and return the data designed as above.

from flask import Flask, jsonify, request



app = Flask(__name__)


@app.route("/")
def root():
    return "Hey There?"


@app.route("/api", methods=["POST"])
def api():
    request_data = request.get_json()
    entity = request_data.get("entity", "")
    if entity:
        resp = query_shareholding(entity)
        data = make_graph_response(resp)
    else:
        data = dict() # tbd
    return jsonify(data)


def parse_nebula_graphd_endpoint():
    ng_endpoints_str = os.environ.get(
        'NG_ENDPOINTS', '127.0.0.1:9669,').split(",")
    ng_endpoints = []
    for endpoint in ng_endpoints_str:
        if endpoint:
            parts = endpoint.split(":")  # we dont consider IPv6 now
            ng_endpoints.append((parts[0], int(parts[1])))
    return ng_endpoints

def query_shareholding(entity):
    query_string = (
        f"USE shareholding; "
        f"MATCH p=(v)-[e:hold_share|:is_branch_of|:reletive_with|:role_as*1..3]-(v2) "
        f"WHERE id(v) IN ['{ entity }'] RETURN p LIMIT 100"
    )
    session = connection_pool.get_session('root', 'nebula')
    resp = session.execute(query_string)
    return resp

And by starting this Flask App instance:

export NG_ENDPOINTS="192.168.8.137:9669"
python3 app.py

 * Serving Flask app 'app' (lazy loading)
 * Environment: production
   WARNING: This is a development server. Do not use it in a production deployment.
   Use a production WSGI server instead.
 * Debug mode: off
[2021-10-13 18:30:17,574]: * Running on all addresses.
   WARNING: This is a development server. Do not use it in a production deployment.
[2021-10-13 18:30:17,574]: * Running on http://192.168.10.14:5000/ (Press CTRL+C to quit)

we could then query the API with cURL like this:

curl --header "Content-Type: application/json" \
     --request POST \
     --data '{"entity": "c_132"}' \
     http://192.168.10.14:5000/api | jq

{
  "nodes": [
    {
      "id": "c_132",
      "name": "\"Chambers LLC\"",
      "tag": "corp"
    },
    {
      "id": "c_245",
      "name": "\"Thompson-King\"",
      "tag": "corp"
    },
    {
      "id": "c_132",
      "name": "\"Chambers LLC\"",
      "tag": "corp"
    },
...
    }
  ],
  "relationships": [
    {
      "edge": "hold_share",
      "properties": "{'share': 0.0}",
      "source": "c_245",
      "target": "c_132"
    {
      "edge": "hold_share",
      "properties": "{'share': 9.0}",
      "source": "p_1767",
      "target": "c_132"
    },
    {
      "edge": "hold_share",
      "properties": "{'share': 11.0}",
      "source": "p_1997",
      "target": "c_132"
    },
...
    },
    {
      "edge": "reletive_with",
      "properties": "{'degree': 51}",
      "source": "p_7283",
      "target": "p_4723"
    }
  ]
}

Upstreams Projects

Flask
vue.js
vue-network-d3
D3.js
Nebula Graph
Nebula-Python

A PoC Corporation Relationship Knowledge Graph System on top of Nebula Graph.

Related tags

Overview

Quick Start

Design Log

data from Backend Side

Data visualization

The data construction in Back End:

The Flask App

Upstreams Projects

Owner

Wey Gu

Official repository for "Intriguing Properties of Vision Transformers" (2021)

Laser device for neutralizing - mosquitoes, weeds and pests

Unsupervised Image-to-Image Translation

An interactive DNN Model deployed on web that predicts the chance of heart failure for a patient with an accuracy of 98%

Neural machine translation between the writings of Shakespeare and modern English using TensorFlow

Graph Convolutional Networks for Temporal Action Localization (ICCV2019)

AdaNet is a lightweight TensorFlow-based framework for automatically learning high-quality models with minimal expert intervention

A PyTorch implementation of "Signed Graph Convolutional Network" (ICDM 2018).

Open standard for machine learning interoperability

Deep Learning for Natural Language Processing SS 2021 (TU Darmstadt)

Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition

ZSL-KG is a general-purpose zero-shot learning framework with a novel transformer graph convolutional network (TrGCN) to learn class representation from common sense knowledge graphs.

Data Preparation, Processing, and Visualization for MoVi Data

Code of the paper "Deep Human Dynamics Prior" in ACM MM 2021.

Creating a custom CNN hypertunned architeture for the Fashion MNIST dataset with Python, Keras and Tensorflow.

TDmatch is a Python library developed to perform matching tasks in three categories:

An implementation of Deep Forest 2021.2.1.

Office source code of paper UniFuse: Unidirectional Fusion for 360$^\circ$ Panorama Depth Estimation

Codebase for the paper titled "Continual learning with local module selection"

Robotics with GPU computing