Text-Clustering

This project groups the news in the provided news dataset to various clusters using unsupervised learning.

Installation Instructions

The steps here are required only if you want to view/run the Jupyter Notebook (i.e. .ipynb and .py files) on your local machine. If not, you can just proceed to Viewing Instructions: To view the HTML version of the clustering results, which does not require any prior setup.

To run the scripts on your local machine (Assumes that you already have an existing anaconda (Windows/Mac) or conda installation (Linux), with Python installed):

1. Launch Anaconda Prompt (Windows/Mac) / Terminal (Linux).
2. Navigate to Text-Clustering folder (Command below assumes that you downloaded and extracted the Text-Clustering folder to your downloads folder).

$ cd downloads 
$ cd Text-Clustering

3. Create a new Python environment and activate it.

$ conda create -n text-clustering python=3.7
$ conda activate text-clustering

4. Install the required dependencies

$ pip install -r requirements.txt

5. Launch Jupyter Notebook

$ jupyter notebook

Viewing Instructions

To view the Jupyter Notebook version of the clustering results:

• From the Home Page of the Jupyter Notebook, navigate to and open Text-Clustering-with-Top2Vec.ipynb.
• Run all the cells in the notebook from top to bottom (Need to execute this step to view the interactive visualizations).

To view the HTML version of the clustering results:

• Open Text-Clustering-with-Top2Vec.html in your browser.

To view the helper scripts used in Text-Clustering-with-Top2Vec.ipynb:

• Navigate to and open top2vec.py.

Files

Text-Clustering-with-Top2Vec.ipynb

• Jupyter Notebook containing the clustering results

Text-Clustering-with-Top2Vec.html

• HTML version of Text-Clustering-with-Top2Vec.ipynb

top2vec.py

• Contains helper classes and functions for clustering the news dataset

requirements.txt

• Contains the list of dependencies required to run Text-Clustering-with-Top2Vec.ipynb and top2vec.py

README.md

• Contains the instructions for viewing the clustering results

news_data.csv

• Provided news dataset for clustering