Abstract:
Online shopping is a big deal nowadays, and many people love it. But there's this common issue – folks often put stuff in their online cart but not buy anything. It's a puzzle for online shops. This thesis dives into understanding why people do this and aims to predict their behavior using data mining. The main goal is to determine what customers might do when visiting an online store. We're using WEKA to analyze data we collected from an online shop. We split the data into two sets – one with all the details and another more focused based on what the computer thinks is important. We then tried seven different methods in WEKA to see which is best at predicting what customers might do. The idea is to find ways to encourage people who leave the website without buying anything to return and maybe make a purchase. The results of this study give helpful tips that can work for various online shops. We suggest rules and decision trees to predict and encourage customer behaviors, especially for those who leave the website early. The chosen method, which is good at predicting stuff, gives a solid foundation for making intelligent decisions in marketing and improving the user experience. This research aims to help businesses use Watson's data as an example to understand better and respond to what customers do when shopping online.
