Customer segmentation using supervised clustering with association rules

مقدم أطروحة جامعية

Bin Lamlih, Nadiyah

مشرف أطروحة جامعية

Harrud, Hamid
Berrado, Abd al-Aziz

أعضاء اللجنة

Smith, Kevin
Bin Nani, Samir

الجامعة

جامعة الأخوين

الكلية

كلية الهندسة و العلوم

القسم الأكاديمي

هندسة البرمجيات

دولة الجامعة

المغرب

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2008

الملخص الإنجليزي

In today’s competitive environment, many organizations are focusing on the notion of customer satisfaction and loyalty.

Customer Relationship Management is becoming a common and important concept in many industries.

CRM is a process companies use to understand their customer groups and respond quickly to shifting customer needs and behaviors.

One of the important approaches in CRM that help firms understand their customers is Customer Segmentation.

Segmentation is partitioning a population of customers into different groups that share similar characteristics.

The Clustering technique is usually used for this task.

Clustering algorithms partition data sets into groups of objects such that the pairwise similarity between objects within the same cluster is higher than those assigned to different clusters.

Since customer data contains categorical values, defining a similarity measure becomes challenging and affects the quality and meaningfulness of the clusters formed.

Furthermore, the curse of dimensionality diminishes the robustness of such measures.

Recent research [1] has resulted in a new non-traditional algorithm for clustering high dimensional categorical data, namely SCAR (Supervised Clustering with Association Rules).

SCAR is robust to the curse of dimensionality, it relies on association rules as an intuitive way to evaluate the similarity between objects and group them.

The proposed algorithm has an optional step for further grouping of the clusters formed initially.

In this project, we introduce an implementation of this new algorithm and use it to segment survey data collected to understand the variables that affect purchase intentions in Moroccan large retailers.

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

عدد الصفحات

106

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-629857