yoheinakajima/instagraph

Converts text input or URL into knowledge graph and displays

Note for non-coders: you can sign up for the waitlist at instagraph.ai.

InstaGraph 🌐

Hello there, adventurous coder! Welcome to InstaGraph, your go-to application for converting text or URLs into insightful knowledge graphs. Curious about the relationships between entities in a complex topic? Feed the text to InstaGraph and voila! A beautiful knowledge graph is at your fingertips.

See example flowcharts generated by InstaGraph here.

Powered by OpenAI's GPT-3.5, this Flask application turns your text into a vividly colored graph, making it easier to visualize relationships between various entities. Enough talking—let's get started!

Author's TL;DR: If you're just looking for how the knowledge graph is generated, check out the function call parameters taking up half of main.py.

Table of Contents 📚

• Features
• Installation
• Usage
• Contributing
• License

Features 🌟

• Dynamic Text to Graph conversion.
• Color-coded graph nodes and edges.
• Responsive design—use it on any device.
• Super-duper user-friendly!

Installation 🛠️

To get started, you'll need Python and pip installed.

1. Clone the repository

git clone https://github.com/yoheinakajima/instagraph.git

2. Navigate to the project directory

cd instagraph

3. Install the required Python packages

pip install -r requirements.txt

4. Set up your OpenAI API Key

Change .env.example to .env

mv .env.example .env

Add your OpenAI API key to .env file:

OPENAI_API_KEY=your-api-key-here

Optional - Set a Graph Database

Use: [--graph neo4j|falkordb] to select the Graph Database driver

• Neo4J

Add Neo4J username, password and URL in the *.env file as well by creating an instance of neo4j.

NEO4J_USERNAME=
NEO4J_PASSWORD=
NEO4J_URI=

• FalkorDB

Add FalkorDB URL in the *.env file as well by creating an instance of FlakorDB.

FALKORDB_URL=

5. Run the Flask app

python main.py [--graph neo4j|falkordb] [--port port] [--debug]

Navigate to http://localhost:8080 to see your app running.

Run as Container

1. Clone the repository

git clone https://github.com/yoheinakajima/instagraph.git

2. Navigate to the project docker directory

cd instagraph/docker

3.1 Run in Dev mode

docker-compose -f docker-compose-dev.yml up # Add -d flag at the end to run in background/daemon mode.

3.2 Run in Prod - Create the docker image

• Using the gunicorn==21.2.0 to run the application in production mode

docker-compose -f docker-compose.yml up --build -d

Usage 🎉

Web Interface

• Open your web browser and navigate to http://localhost:8080.
• Type your text or paste a URL in the input box.
• Click "Submit" and wait for the magic to happen!

API Endpoints

1. GET Response Data: /get_response_data

   • Method: POST
   • Data Params: {"user_input": "Your text here"}
   • Response: GPT-3.5 processed data

2. GET Graph Data: /get_graph_data

   • Method: POST
   • Response: Graph Data

3. GET History Data: /get_graph_history

   • Method: GET
   • Response: Graph Data

Contributing 🤝

Best way to chat with me is on Twitter at @yoheinakajima. I usually only code on the weekends or at night, and in pretty small chunks. I have lots ideas on what I want to add here, but obviously this would move faster with everyone. Not sure I can manage Github well given my time constraints, so please reach out if you want to help me run the Github. Now, here are a few ideas on what I think we should add based on comments...

• Store knowledge graph (thx @tomasonjo! 9/13/23)
• Pull knowledge graph from storage (thx @tomasonjo! 9/13/23)
• Show history
• Ability to combine two graphs
• Ability to combine 2+ graphs from history
• Ability to expand on a graph (thx @tomasonjo! 9/13/23)
• Ability to expand on graph from specific nodes
• Fuzzy matching of nodes for combining graphs (vector match + LLM confirmation)

There are a lot of "build a chart" tools out there, so instead of doing user account and custom charts, it sounds more fun for me to work on building the largest knowledge graph ever...

Before creating an Issue please refer the ISSUE_TEMPLATE provided.

License 📝

MIT License. See LICENSE for more information.

---

Enjoy using InstaGraph! 🎉

instagraph's Issues

[Feature Improvements] Gitignore file update

We should have a gitignore file to protect from unwanted file push on GitHub.

[BUG]Page request twice, resulting in generating two charts

I noticed on the console that the conversation sent two/get_response_data requests, resulting in the chart being updated for the second time.
The second question is, I found that some of the answers I received were too simple and not practical, such as asking: What are the directions for thinking? Its answer is direction one and direction two, rather than actual type names, such as life or work

[BUG] Error fetching graph history

Describe the bug
When entering http://localhost:8080 a message appears saying the following statement: "Graph History
Error fetching graph history: Error: Network response was not ok INTERNAL SERVER ERROR". Also the submit button does nothing.

To Reproduce
Steps to reproduce the behavior:

1. Clone the repository.
2. Install the requirements.
3. Prepare the environment by adding your Openai key and changing the name of the env file to .env.
4. Run main.py.
5. Enter localhost to see the error.

Expected behavior
A working text input field and Submit button.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [Windows 10]
• Browser [Chrome and Firefox]
• Version [latest]

Additional context
Add any other context about the problem here.

[Question] Scrape text from website function never used?

Following function is never used:

def scrape_text_from_url(url):
	response = requests.get(url)
    if response.status_code != 200:
        return "Error: Could not retrieve content from URL."
    soup = BeautifulSoup(response.text, "html.parser")
    paragraphs = soup.find_all("p")
    text = " ".join([p.get_text() for p in paragraphs])
    print("web scrape done")
    return text`

[Question] should it update neo4j graph?

I'm getting graph in instagraph but in neo4j database is empty,
should it be like that ? I though it will also update neo4j

Why am I getting error code 429 everytime ""POST /get_response_data HTTP/1.1" 429"

I set up everything from installation to Open AI API Key and neo4j credentials however every time I try to generate graph I get nothing but this error code in terminal ""POST /get_response_data HTTP/1.1" 429"
I have some idea that it has to do something API rate limit but but this is my first time using Open AI API Keys.
here is screenshot:

[Question] Graphviz or Cytoscape

I can't seem to find where graphviz is used, is it old code?

Thank you!

[BUG] Not generating graphs from URL

Describe the bug
Generates graphs from a text prompt but no graph is generated from a URL input.

To Reproduce
Steps to reproduce the behavior:

1. Clone the repository.
2. Install the requirements.
3. Prepare the environment by adding your Openai key and changing the name of the env file to .env.
4. Run main.py.
5. Open localhost and enter a URL (e.g. https://aws.amazon.com/what-is/quantum-computing/)
6. Wait for the loading bar to finish, no graph is generated

Expected behaviour
Generation of a graph from the URL.

Screenshots

Desktop (please complete the following information):

• OS: macOS
• Browser firefox
• Version [Latest]

Free users go beyond rate limits

Free users can only do 3 requests per minute, by default when submitting a request the app doesn't wait and the quote is exceeded, and a stack trace is shown on the console launching the flask app.

A solution for free users would be to wait 20 or 30 seconds between each request towards openai api.

Not able to setup the environment to run the code

After completing all the steps in README file, and troubleshooting some things by my self. I got the following error while running "python main.py" in terminal

NotOpenSSLWarning: urllib3 v2.0 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See: https://github.com/urllib3/urllib3/issues/3020
  warnings.warn(
Traceback (most recent call last):
  File "/Users/dhandeepsingh/Desktop/instagraph/main.py", line 5, in <module>
    from bs4 import BeautifulSoup
ModuleNotFoundError: No module named 'bs4'

How to fix this issue??

[Suggestion] To discuss, learn, build, contribute together!

I find this project much interesting, having a good place to discuss would add up value to the community. Consider creating a place to discuss.

1. Github Discussion
2. Slack
3. Discord Channel

problem installing the requirement neo4j , i keep getting this error

Collecting neo4j==5.12.0
Downloading neo4j-5.12.0.tar.gz (190 kB)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 190.9/190.9 KB 2.0 MB/s eta 0:00:00
Installing build dependencies ... done
Getting requirements to build wheel ... done
Installing backend dependencies ... done
Preparing metadata (pyproject.toml) ... done
WARNING: Generating metadata for package neo4j produced metadata for project name unknown. Fix your #egg=neo4j fragments.
Discarding https://files.pythonhosted.org/packages/19/11/9af80c8621ba433039bf47acd7b037998138516eb86fdb9ba492c49ee844/neo4j-5.12.0.tar.gz#sha256=00a776a687267150f9e1950017316b751cf63db7b734a699b1405ac20fd4a731 (from https://pypi.org/simple/neo4j/) (requires-python:>=3.7): Requested unknown from https://files.pythonhosted.org/packages/19/11/9af80c8621ba433039bf47acd7b037998138516eb86fdb9ba492c49ee844/neo4j-5.12.0.tar.gz#sha256=00a776a687267150f9e1950017316b751cf63db7b734a699b1405ac20fd4a731 (from -r requirements.txt (line 6)) has inconsistent name: filename has 'neo4j', but metadata has 'unknown'
ERROR: Could not find a version that satisfies the requirement neo4j==5.12.0 (from versions: 1.7.0b1, 1.7.0b2, 1.7.0b3, 1.7.0b4, 1.7.0rc1, 1.7.0rc2, 1.7.0, 1.7.1, 1.7.2, 1.7.3, 1.7.4, 1.7.5, 1.7.6, 4.0.0a1, 4.0.0a2, 4.0.0a3, 4.0.0a4, 4.0.0b1, 4.0.0b2, 4.0.0rc1, 4.0.0, 4.0.1, 4.0.2, 4.0.3, 4.1.0rc1, 4.1.0rc2, 4.1.0, 4.1.1, 4.1.2, 4.1.3, 4.2.0a1, 4.2.0, 4.2.1, 4.3.0a1, 4.3.0b1, 4.3.0rc1, 4.3.0, 4.3.1, 4.3.2, 4.3.3, 4.3.4, 4.3.5, 4.3.6, 4.3.7, 4.3.8, 4.3.9, 4.4.0a1, 4.4.0b1, 4.4.0, 4.4.1, 4.4.2, 4.4.3, 4.4.4, 4.4.5, 4.4.6, 4.4.7, 4.4.8, 4.4.9, 4.4.10, 4.4.11, 5.0.0a1, 5.0.0a2, 5.0.0, 5.0.1, 5.1.0, 5.2.0, 5.2.1, 5.3.0, 5.4.0, 5.5.0, 5.6.0, 5.7.0, 5.8.0, 5.8.1, 5.9.0, 5.10.0, 5.11.0, 5.12.0)
ERROR: No matching distribution found for neo4j==5.12.0

Error in installing

ERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.
transformers 2.1.1 requires sentencepiece, which is not installed.
anaconda-cloud-auth 0.1.4 requires pydantic<2.0, but you have pydantic 2.4.2 which is incompatible.
python-lsp-black 1.2.1 requires black>=22.3.0, but you have black 0.0 which is incompatible.

I get the above message in windows powershell when using this command;
pip install -r requirements.txt

What would be the next step to move forward with the install?

[feature enhancement] .env file create

I have purporsal to make a .env file, so any other can easily store the OpenAI api key easily and run in their local machine.

Create the public docker image using the docker registries like dockerhub/gcr

@yoheinakajima Can we create a organization with "instagraph"
Create the docker image with the same. Like other technologies have the dockerhub account and image with the organization name.

Can the generated graph be manually fixed?

Once a graph is generated, what if there are some connections or nodes be fixed manually?

Template html seems broken

i get this instead of a graph

looking at the index.html it seems rekt (some text meant to be in script tags are just hanging out below the html closing tag

Graph

console has been displaying the following - 127.0.0.1 - - [16/Sep/2023 11:04:40] "POST /get_response_data HTTP/1.1" 200 but the graph is not being displayed at http://127.0.0.1:8080

UI upgrade ideas

Here are some ideas for an upgrade to the UI @yoheinakajima

[Feature Request] Is it possible to display content in real-time?

At present, the page is displayed after waiting for all messages to reply. Can we continuously display the page content based on the streaming output, so that we don't need to wait for a long time

[Question] how to eliminate creation of duplicated nodes

If I ask the same question few times, it will create nodes with same name
if I will use similar world like DB or Database or Azure SQLDB or Azure SQL DB, it will also create duplicated.
where I can change it? maybe idea for feature.

[Question]what should I do to deal with this problem when I'm installing.

Add support for other Graph Databases

Please add a generic API so other Graph Databases can be support too.

[Enhancement] Code of conduct

Contribution.MD is importance to other to follow the code of ethics.

Bigger UI in browser

I have a vary narrow screen in my browser (after opening localhost 8080) How can I increase the size so I can see the graph much more clear?

[Feature Request] Support multiple LLMs using litellm

Hi Team,

Not sure if you can support different LLMs in your app?

https://github.com/BerriAI/litellm

Thanks.

[Feature] Creating templates for different category

create template for bug_report, Feature, Question.

Add ability to support Embeddings, Vector Storage and Retrieval Augmentation Methods [Feature Request]

It would be great to have plug-and-play support for Embeddings, Vector Storage and Retrieval Augmentation
That way the graphs can have a lot more context

Further on that path - adding embedding models so that the user can upload or link as much data as they like
Entire coursework for example

I and my team would be happy to work on this and contribute however possible
Thank You

Error when running Flask application

Describe the bug
Error when running Flask application
新建文本文档.txt

(myenv) D:\my projects\instagraph>python main.py
File "D:\my projects\instagraph\main.py", line 262
match graph.lower():
^
SyntaxError: invalid syntax

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: [winows]
• Browser [chrome]
• Version [10]

Additional context
Add any other context about the problem here.

[Feature Request] Contribution.MD file

Contribution docs set the rule, for Open source contribution.

[Feature Request]Hope to support Azure OpenAI service

Hope to support Azure OpenAI service

[Feature] Docker file is missing

Many prefer to run the code on docker so they didn't require to manually install the requirements.

Assign me I will complete.

[Question] what is `response_model` ?

Hi! can someone explain what the response_model param used here is:

https://github.com/yoheinakajima/instagraph/blob/main/main.py#L144C18-L144C18

this seems to format the response in exactly the data structure of the models provided.
Is this an openAI thing, or part of fast API? I can't find any docs on openAI's site.

https://platform.openai.com/docs/api-reference/chat

If its an openAI thing perhaps its also magically reading the heredoc of the python classes?

I'm actually trying to use the API via typescript, so I was wondering what the parallel might be.

[BUG] NameError: name 'user_plan' is not defined

Describe the bug
I'm getting the following error when trying to open the link

$ python3 main.py
 * Serving Flask app 'main'
 * Debug mode: off
WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on all addresses (0.0.0.0)
 * Running on http://127.0.0.1:8080
 * Running on http://192.168.1.132:8080
Press CTRL+C to quit
[2023-09-22 03:59:57,400] ERROR in app: Exception on / [GET]
Traceback (most recent call last):
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 2190, in wsgi_app
    response = self.full_dispatch_request()
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 1487, in full_dispatch_request
    return self.finalize_request(rv)
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 1508, in finalize_request
    response = self.process_response(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 2002, in process_response
    response = self.ensure_sync(func)(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 66, in add_header
    if check_if_free_plan():
       ^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 61, in check_if_free_plan
    return user_plan == 'free'
           ^^^^^^^^^
NameError: name 'user_plan' is not defined
[2023-09-22 03:59:57,400] ERROR in app: Request finalizing failed with an error while handling an error
Traceback (most recent call last):
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 2190, in wsgi_app
    response = self.full_dispatch_request()
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 1487, in full_dispatch_request
    return self.finalize_request(rv)
           ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 1508, in finalize_request
    response = self.process_response(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 2002, in process_response
    response = self.ensure_sync(func)(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 66, in add_header
    if check_if_free_plan():
       ^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 61, in check_if_free_plan
    return user_plan == 'free'
           ^^^^^^^^^
NameError: name 'user_plan' is not defined

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 1508, in finalize_request
    response = self.process_response(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/.local/lib/python3.11/site-packages/flask/app.py", line 2002, in process_response
    response = self.ensure_sync(func)(response)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 66, in add_header
    if check_if_free_plan():
       ^^^^^^^^^^^^^^^^^^^^
  File "/home/guy/workspace/instagraph/main.py", line 61, in check_if_free_plan
    return user_plan == 'free'
           ^^^^^^^^^
NameError: name 'user_plan' is not defined
127.0.0.1 - - [22/Sep/2023 03:59:57] "GET / HTTP/1.1" 500 -

To Reproduce
Steps to reproduce the behavior:

1. Run python3 main.py
2. Click on one of the links

Desktop (please complete the following information):

• OS: Ubuntu 23.04
• Browser Firefox
• Version 117.01

Additional context
Add any other context about the problem here.

Is it possible to download the knowledge graph locally on the web page?[Feature Request]

Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]

Describe the solution you'd like
A clear and concise description of what you want to happen.

Additional context
Add any other context or screenshots about the feature request here.

Enhancement (drafted) - extend graph to allow latex equations in popups

https://chat.openai.com/share/ab2b5332-30c5-4900-b607-be4c89da3f60