ArctiPy

An app built to simplify and condense NLP tasks into one simple yet powerful Interface.

Setup

Clone the Repository

To use this app, simple clone the repository onto your local system, navigate into the directory of the cloned repository and run the following commands in your favourite Python Virtual Environment!

pip install -r requirements.txt

streamlit run app.py

Docker

If you do not wish to set up an Environment to run the app, you may choose to run the app using Docker instead! We have created pre-made Docker images hosted on Github Packages for you to use. To do so, simply install Docker on the target system and run the following commands on Terminal or Powershell:

docker pull ghcr.io/asdfghjkxd/app:main

docker run -it -p 5000:8501 --name news ghcr.io/asdfghjkxd/app:main

The created Docker Container can then be accessed through localhost on Port 5000!

If Command Lines are not your thing, you can do the same using the Docker Desktop GUI! Just follow the steps below to set up the Container:

• Open up Terminal or Powershell and key in the command docker pull ghcr.io/asdfghjkxd/app:main word for word (we promise this is the only Command Line step in the entire process!)
• After the image has been downloaded, open up the Docker Desktop app
• Click on the Images tab on the sidebar and find the image you have pulled in the above step
• Click on the Run button
• Click on the Optional Settings button
• Enter in the variables you want to use in the fields that appear
  • Container Name: Enter a suitable name for the Docker Container you are about to create
  • Ports > Local Host: Key in a suitable and available port on your device; this port will be used to access the app through your device (corresponds to [YOUR_PORT_ABOVE] below)
  • Volumes > Host Path: Path to a folder on your device to store any saved date from the Docker Container to allow persistence of data created in the Container (optional, as files are not passed over to the Docker Host through the persisted folder mounted onto the Docker Container)
  • Volumes > Container Path: Path to a folder on the Container, should be in the format /usr/[PATH] (optional, as files are not passed over to the Docker Host through the persisted folder mounted onto the Docker Container)
  • Click on Run and navigate to the localhost address localhost:[YOUR_PORT_ABOVE] on your web browser

Web App

If you do not wish to set up your system at all, and you do not mind using the app through the Internet, you may use the app on the website https://share.streamlit.io/asdfghjkxd/articpy/main/app.py! You can quickly access the website by scanning the QR Code below too!

Do note that the performance of the app is restricted to the resources available for the container used to host the app online. If you wish to use a GPU, or if your workflow requires greater computing power, kindly use the other methods outlined above to run the app.

Usage

There are 3 main modules in this app, each performing an important step of the way for NLP analysis.

Load, Clean and Analyse

This module is the first module you will be using to preprocess your data before conducting further analysis on it.

There are 3 things this module can do: Data Cleaning, Data Modification and Data Query.

• Data Cleaning: Cleans and preprocess your raw data for futher analysis
• Data Modification: Allows you to extract the mentions of countries within the documents or allow you to modify your data inplace
• Data Query: Query for certain strings in the document

Document-Term Matrix

This module will allow you to create Document-Term Matrix and Word Use Frequency Data.

NLP Toolkit

This module will allow you to conduct advanced NLP analyses on your processed dataset. You are able to perform the following tasks on your dataset:

• Topic Modelling: Models your documents using statistical methods
• Topic Classification: Classify your documents using labels you define
• Sentiment Analysis: Analyse the sentiment of your doucments using statistical methods
• Word Cloud Creation: Creates a word cloud visualisation of the entire collection of documents
• Named Entity Recognition: Conducts NER on your documents
• Position of Speech Tagging: COnducts POS on your documents
• Summary: Summarises each document in the collection to your specified parameters