RISKOUT - Risk Management Platform for Military

--- ### Quick Links

:book: Table of Contents

➤ Intro
➤ Features

Risk Dashboard
Risk Detection
Report Generator

➤ Getting Started
➤ Prequisites
➤ Techniques Used
➤ Installation Process
➤ Team Information
➤ Copyleft / End User License

:monocle_face: Intro

There are numerous threats to military. There are external threats such as spies, hackers, terrorists, but the actual threats of current well-established military are internal threats such as leaked secrets, fake news, and malicious posts. So how does military identify and manage these?

Currently, military searches and captures leaked documents and fake news 24/7 and manually cuts news from newspapers. The collected data is then read and organized by soldiers into reports, finally handing them over to the response team. Due to the long and complex process, human errors or delayed responses may occur.

So we thought. Let's create an all-in-one platform that can automatically identify and manage malicious risks such as leaked secrets and fake news. That's when RISKOUT was born.

:plate_with_cutlery: Features

3 Main Features are:

💀 Risk Dashboard: Dashboard that visualizes public sentiment and the current state of media coverage
😤 Risk Detection : Threat detection page that automatically identifies and analyzes malicious posts, such as leaks of military secrets and fake articles
📰 Report Generator : Generates customizable threat reports with just a few clicks

Risk Dashboard

Keywords for Today

You can find more details about "Keywords for Today" feature here.

A word cloud that visualizes the most frequently mentioned words based on various articles, news, and various online communities.

words

Emotion Recognition Chart

You can find more details about "Emotion Recognition Chart" feature here.

Charts that analyze the sentiment of public opinion based on various social media and community sites, categorizing it into positive, neutral, and negative.

emopies

Today's Top Trend

You can find more details about "Today's Top Trend" feature here.

Selects the most mentioned articles of the day and uses FactCheck to classify and display them as likely true, neutral, or false.

trend

Events by Country

You can find more details about "Events by Country" feature here.

A map that analyzes international articles to show event traffic by country.

events

Article Variation

You can find more details about "Article Variation" feature here.

A chart that visualizes changes in the volume of articles by comparing recent article quantities.

num_articles

Risk Detection

You can find more details about "Risk Detection" feature here.

Using artificial intelligence, it automatically analyzes and detects malicious posts such as leaks of confidential information and fake articles. It then provides summarized content and sources to enable a quick response.

detect

NER Filter

You can find more details about "NER Filter" feature here.

Utilizes Named Entity Recognition technology to extract types such as people, organizations, and dates, offering search filters to enable more detailed analysis.

ner

Automated Report Generation

You can find more details about "Automated Report Generation" feature here.

Automatically organizes and summarizes the threats identified into a report format with just a few clicks. The generated report can be exported as a PDF.

report

report_full

:zap: Getting Started

After loggin in:

Congratulations! You joined RISKOUT!.

That's all you need to get started! 🎉

📺 Full video: https://www.youtube.com/watch?v=Lwg-OQIIvGA

:fork_and_knife: Prerequisites

:earth_asia: Browser

Chrome	Internet Explorer	Edge	Safari	Firefox
Yes	11+	Yes	Yes	Yes

💾 Versions

Pytorch	React	Django	🍃 Mongo DB	🐳 Docker	Ⓜ️ MUI
1.9.0+	17.0.2+	3.0.7+	4.4+	20.10.x+	5.0.1+

🧱 Technique Used

techstack

AI

Colabfor AI model training:
- KoBERT — for sentiment analysis, fake news detection, and report summarization.
- DistilKoBERT — for Named Entity Recognition (NER).
Datasets used:
- Naver-nsmc — Datasets used for the sentiment analysis model.
- Dacon document summarization AI — Datasets used for Korean document extraction and summarization.
- SNU Factcheck — Datasets used for fake news detection.
- Naver NLP Challenge 2018 — Datasets used for Named Entity Recognition (NER).
Pytorch Libraries used for deep learning build.
- Transformers — Provided architecture for NLP models.
- FastAPI — Implementing AI functionality through APIs.

Backend

DRF for backend development:
- Mongo DB — for database development.
Beautiful Soup Crawling using:
- Crawler — For extracting language data from various open forums, social media platforms, and news sites.

Frontend

React for frontend development:
- MUI — Utilizing the MUI (Material UI) component library.
- React router — Used for component navigation.
Recoilfor state management in React.
- Atom — for separating component state units.
- Selector — for generating dynamic data dependent on Atoms.

:file_folder: Installation Process

First, download node.js, yarn, docker, and docker-compose. Ensure that node.js is version 14.x or higher.

Clone the project.

git clone https://github.com/osamhack2021/ai_web_RISKOUT_BTS

Create the Secret files.

For information on creating Secret files, please refer here.

Build and run the project.

./run.sh

Access the project athttp://localhost:8002.

You can now start using it! 🎉

💁🏻‍♀️💁🏻‍♂️ Team Information

Profile	Name	Role	Github	E-mail
	Minseok Lee (Team Leader)	Product Manager, Full-stack Developer
	Junghwan Cho	AI Developer
	Myeonggeun Seo	Frontend Engineer
	Taewon Kim	Backend Engineer
	Wonbin Lee	Frontend Engineer
	Yongjun Park	Backend Engineer
	Jongchan Seo	Frontend Engineer