Application of Data Mining in e-Commerce

Application of Data Mining in e-Commerce

Mohamed Chajri (Sultan Moulay Slimane University, Morocco) and Mohamed Fakir (Sultan Moulay Slimane University, Morocco)
DOI: 10.4018/978-1-4666-8619-9.ch015


The web in recent years has been a big trend, which helped make it a source of information and essential in the various fields of research, in particular, the commercial area that represents the e-commerce (electronic commerce). However, the competition in the e-commerce sites is very tight. This has pushed companies to conserve and retain customers rather than seeking to expand its market share by conquering politically. These requirements have introduced the extraction of knowledge from data in e-commerce sites, using data mining techniques. This article will be an introduction to the concept of data mining, a definition of economic concepts related to e-commerce, and the authors' approach to the application of data mining techniques in e-commerce.
Chapter Preview

2. Data Mining Tasks

Data mining tasks can be divided into 5 parts:

2.1. Description

This is often one of the first tasks required of a data mining tool asked to describe a complex data base. This often creates an additional operation to provide explanations. For example, a customer calls can be summarized in total minutes, total number of calls, and so on.

2.2. Classification

The classification determines the class of an object based on its attributes. A set of objects is given as the training set. Each object is represented by a vector of attributes followed by its class. The function or classification model is built by analyzing the relationship between the attributes and classes in the training set. This function or model can then classify objects in the future.

2.3. Clustering

Segmentation is to extract previously unrecognized groups, having the same characteristics, called clusters. Alternatively it is segmenting a heterogeneous population into homogeneous populations. Unlike classification, subpopulations are not predetermined.

2.4. Associations Rules

This problem is to find the connection between objects, this relationship called rule associations. An association rule indicates that the appearance of a set of objects in a database is strongly related to the appearance of a set of other objects.

2.5. Prediction

The prediction is similar to classification except for with the prediction the results are in the future. Examples of tasks applied to marketing forecast: “Predicting the commercial value of a stock three months in the future”.


3. E-Commerce

E-commerce or electronic commerce includes all commercial transactions taking place remotely through electronic and digital interfaces and essentially involves commercial transactions taking place on the Internet from different types of terminals (computers, tablets, smart phones, consoles).

Although everyone believes that e-commerce is a new technological innovation, the term e-commerce is not entirely new. Indeed trade existed since the 60s thanks primarily to standard EDI (electronic data interchange). But the trend is related to the evolution of the internet.

Complete Chapter List

Search this Book: