Contributors: Nico Van den Hooff, Ting Zhe (TZ) Yan, Arijeet Chatterjee

This repository contains our work for our project focused on applying machine learning classification models to e-commerce data. Specifically, we built a Random Forest classification model in order to determine if a website visitor will complete a purchase or not.

The research question that we attempted to answer with our analysis is stated as follows:

Given clickstream and session data of a user who visits an e-commerce website, can we predict whether or not that visitor will make a purchase?

Nowadays, it is common for companies to sell their products online, with little to no physical presence such as a traditional brick and mortar store. Answering this question is critical for these types of companies in order to ensure that they are able to remain profitable. This information can be used to nudge a potential customer in real-time to complete an online purchase, increasing overall purchase conversion rates. Examples of nudges include highlighting popular products through social proof, and exit intent overlay on webpages.

The dataset used in our analysis was obtained from the UC Irvine Machine Learning Repository, a popular website with hundreds of datasets available for analysis. The creators of the dataset are C. Sakar and Yomi Kastro, and the original dataset can be obtained at this link.

Each row in the dataset contains a feature vector that contains data corresponding to a visit "session" (period of time spent) of a user on an e-commerce website. The dataset was specifically formed so that each session would belong to a unique user over a 1-year period. The total number of sessions in the dataset is 12,330.

Examples of features included in the dataset are:

• Duration (time) spent on Administrative, Informational, or Product related sections of the website.
• Data from Google Analytics such as bounce rate, exit rate, and page value.
• What month the session took place it, and whether or not the day of the session falls on a weekend.
• The operating system and browser used by the website visitor.

The target feature and class label in the dataset is called Revenue and contains either a True or False value, which correspond to whether or not the user made a purchase on the website during their visit, respectively. It is worth noting that the dataset is unbalanced, as 85% of the sessions contain a False class label, with the remaining 15% containing a True label.

Our analysis includes a detailed table that describes each feature in detail.

We have published a copy of our final analysis in the form of a Jupyter Book

To replicate our analysis, you will first need to clone our repo. You can do so with the following command line (terminal) commands.

# clone our repo
git clone https://github.com/UBC-MDS/online-shoppers-purchasing-intention.git

# change working directory to our repos root directory
cd online-shoppers-purchasing-intention

You can automatically replicate our analysis by using one of the following two methods:

First, install Docker.

Then, run the following command to reset our repository to a clean state, with no intermediate or results files:

docker run --rm -v /$(pwd):/home/analysis nicovandenhooff/online-shoppers-purchasing-intention:latest make -C /home/analysis clean

Finally, run the following command to replicate our analysis (note: this will take some time):

docker run --rm -v /$(pwd):/home/analysis nicovandenhooff/online-shoppers-purchasing-intention:latest make -C /home/analysis all

First, create and activate the required virtual environment with conda at the command line as follows:

conda env create -f environment.yaml
conda activate project_env

Then, run the following command at the command line (terminal) to reset our repository to a clean state, with no intermediate or results files:

make clean

Finally, run the following command to replicate our analysis (note: this will take some time):

make all

Python 3.9 and Python packages

• altair=4.1.0
• numpy=1.21.2
• pandas=1.3.3
• scikit-learn=1.0
• scipy=1.7.1
• pip>=20
• ipykernel=6.5.0
• docopt=0.6.2
• xgboost=1.5.0
• altair_saver
• altair-data-server==0.4.1
• jupyter-book==0.12.1

The source code for the site is licensed under the MIT license, which you can find here.

Sakar, C., and Kasto, Yomi. 2018. “UCI Machine Learning Repository.” University of California, Irvine, School of Information; Computer Sciences. https://archive-beta.ics.uci.edu/.