ML-Wikipedia-Runner

Machine Learning model that predicts which page best to choose to reach the goal page.

This is a final project for "Introduction to Machine Learning by Università di Macerata" course. Created by Dmytro Zakharov, Daniil Korenkov, Kvaratskheliia David.

Project presentation

You can download it here or from the presentation folder.

What is included in the project?

• internal/dataset_generator — logic responsible for generating the dataset. Main function here is web_clicker.py which launches the browser from the web driver, open the Six Degrees of Wikipedia website and mines the data. In the inputs file we included the list of words that is used for the model and forming the input. output is a place where the generated table is stored.
• internal/model — place where we trained and saved our model.
• interface — scripts that launches the colorful CLI for interacting with a model.

How to launch and which functions are included?

From the project root folder type in:

• python3 main.py -m generate to launch dataset miner
• python3 main.py -m six-degrees-chart to show distances distribution from the Six Degrees of Wikipedia website
• python3 main.py -m generator-chart to show distances distribution for our dataset
• python3 main.py -m generate-features to convert generated dataset to one needed for the model (with features)