Customer Behavior Analysis in E commerce using RNN

Market segmentation allows companies to get a better understanding of their customer's needs and wants. And then the companies can provide recommendations to products the customers are most likely to purchase, in order to boost sales. Segmentation helps marketers to be more efficient in terms of tim

Project Title

Customer Behavior Analysis in E commerce using RNN

Project Area of Specialization

Artificial Intelligence

Project Summary

Market segmentation allows companies to get a better understanding of their customer's needs and wants. And then the companies can provide recommendations to products the customers are most likely to purchase, in order to boost sales. Segmentation helps marketers to be more efficient in terms of time, money, and other resources. The objective of this project is to provide an e-commerce application that allows customers to get unique personalized recommendations based on market segmentation. Market segmentation will be carried out using machine learning techniques. 

Project Objectives

As businesses move to sell online, the factor of customer experience is vital. As keeping customers is less expensive than making new ones. Recommendation systems that give personalized recommendations to customers, help us in keeping the experience intact. This not only helps the customers but also the retailers. By recommending which products the customers are most likely to buy. Feature engineering for such recommendation systems can be expensive and need a lot of data to start recommending.  

Project Implementation Method

Using RNN-LSTM we will build a model to use the RFMC market segmentation model to make predictions to what the consumer may buy and also take into account the trends throughout the year. We intend to use React for our front end. Node for the Backend and a SQL database.

Benefits of the Project

The task of recommending products to customers can be formulated as a rating prediction problem or as a product ranking problem. Generally collaborative, content-based and hybrid is the methodology used for this type of problem. However several studies show that rating prediction is not efficient in generating top-n recommendations.  Applying RNNs directly to sequences of consumer actions yields the same or higher prediction accuracy than vector-based methods like logistic regression. Unlike the latter, the application of RNNs comes without the need for extensive feature engineering. Adding a model to an e-commerce site will improve customer experience.  The recommendation system can be used for any product range. This enables small businesses to compete with larger online retailers who employ much more complex recommendation systems. But with RNN we get recommendation results that take time into account. And costs much less as a data scientist is not required to keep working with the crafted features, testing optimality of handcrafted features is an on-going process, which makes it hard to tell if the feature set is optimal or not. In some cases computing the hand-crafted features can lead to an expensive preprocessing of the data. Our model will use Recency, Frequency, and Monetary as features to recommend products to customers.

Technical Details of Final Deliverable

For performance comparison, we will be comparing our result to this paper. A Recurrent Neural Network Approach Hojjat Salehinejad and Shahryar Rahnamayan, Senior Member, IEEE Department of Electrical, Computer, and Software Engineering University of Ontario Institute of Technology Oshawa, Ontario, Canada this is the only paper to use RNN for commercial use as RNN has previously been only used for NLP based problems.

Final Deliverable of the Project

Software System

Core Industry

IT

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com