pocinnovation/discover

An app to generate knowledge maps from a single keyword.

If you see this section, you've just created a repository using PoC Innovation's Open-Source project template. Check the getting started guide.

Discovery

An app to generate knowledge maps from a single keyword.

How does it work?

[Explain how this project is working]

Getting Started

Installation

[Explain how to install all of the project's dependencies]

Quickstart

[Explain how to run this project]

Usage

[Explain how to use this project]

Get involved

You're invited to join this project ! Check out the contributing guide.

If you're interested in how the project is organized at a higher level, please contact the current project manager.

Our PoC team ❤️

Developers

Axel ZENINE	Alex GHERASIE	Thibaut DEGODENNE

Manager

Clément LOEUILLET

Organization

🚀 Don't hesitate to follow us on our different networks, and put a star 🌟 on PoC's repositories

Made with ❤️ by PoC

discover's Issues

[Feature]: Research about graph database & start implementation

Description

You need to research which database to implement and give the right and wrong arguments. Then you should start implementing your chosen solution.

Expected behaviour

/

Environment

/

Proposed solution

Check this: https://www.tigergraph.com/

[Feature]: Implement a GNN model

Description

You need to implement a GNN model and start training it with data from a public dataset and then with the project data.

Expected behaviour

/

Environment

/

Proposed solution

Check this:

• https://paperswithcode.com/dataset/wiki-cs
• https://towardsdatascience.com/how-to-create-a-graph-neural-network-in-python-61fd9b83b54e

[Feature]: Present your research

Description

You need to do some research around GNN to start planning for next week's model implementation.

You should also prepare questions to ask Maxime Labonne.

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Figma sketchs

Description

Creation of a mock-up for the realization of the application

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Workshops

Description

The team must to do this workshops:

• Data science
• Decision tree
• Neural network

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Implement base algorithm for generating graph using the wikipedia API

Description

Generate a graph based on input string using the wikipedia api.

Expected behaviour

/

Environment

/

Proposed solution

/

[Feature]: Research on / implement embeddings

Description

Do some research on embeddings like word2vec to see if / how it can be used for our project.

Expected behaviour

• Ideally a proof of concept on a particular field.
• Ability to target precise information on a topic (like how in word2vec, king - man = queen)
• Decide if this technique is good for our project

Environment

/

Proposed solution

• https://pytorch.org/docs/stable/generated/torch.nn.Embedding.html
• https://www.youtube.com/watch?v=gQddtTdmG_8

[Feature]: Research on the addition of vectors

Description

You need to do more research on vector addition and check if it is feasible in the project.

The objective of this is to adapt the desired graph to the added vectors

Expected behaviour

/

Environment

/

Proposed solution

No response

[Feature]: Implementation fo hydra

Description

You will have to implement hydra for the structure of your model.

Check this:

• https://hydra.ai/
• https://github.com/ORNL/HydraGNN

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Contributing

Description

Description

Develop your Open Source skills 🎉

Expected behaviour

Here is what you have to do :

• Create a project from PoC's project template and play with :

  • Milestones
  • Issues
  • Pull Requests
  • Releases
  • Labels
  • GitHub project

• Update the README.MD file and the contributing guide

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: First data and formatting

Description

You need to start collecting data with the wikipedia api and find a standard formatting.
Work with your teammates for data.

Expected behaviour

/

Environment

/

Proposed solution

Check this:

• https://pypi.org/project/Wikipedia-API/
• https://arxiv.org/pdf/2007.02901.pdf

[Feature]: Improve the GNN notebook

Description

You have to improve the notebook of the GNN model.

Expected behaviour

/

Environment

/

Proposed solution

No response

[Feature]: Custom your Notion space

Description

The team must to do:

• Discover your dedicated space
• Custom your space
• Add your resources

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Wikipedia Application

Description

You need to create an application using the wikipedia api to retrieve the following information from a page name:

• Page exist
• URL
• Primary informations
• Summary
• Text

You can use HuggingFace to deploy a Gradio or Streamlit application.

You can check this:

• https://pypi.org/project/Wikipedia-API/
• https://huggingface.co/
• https://gradio.app/
• https://streamlit.io/

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response

[Feature]: Resolve the challenge

Description

The team must to resolve a DAWG challenge available on this URL: https://www.codingame.com/training/hard/dawg

Expected behaviour

This should happened!

Environment

• OS and version

Proposed solution

No response