Change search
ReferencesLink to record
Permanent link

Direct link
Response modeling in direct marketing: a data mining based approach for target selection
2008 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Identifying customers who are more likely to respond to a product offering is an important issue in direct marketing. In direct marketing, data mining has been used extensively to identify potential customers for a new product (target selection). Using historical purchase data, a predictive response model with data mining techniques is developed to predict a probability that a customer is going to respond to a promotion or an offer. The purpose of this thesis is to identify the Parsian bank customers who are more likely to respond positively to a new product offering. To reach this purpose a predictive response model using customer historical purchase data is build with data mining techniques. Response modeling procedure consists of several steps. In building a response model one has to deal with some issues, such as: constructing all purchase behavior variables (RFM variables), determining the inputs to the model (feature selection) and class imbalance problem. The purpose of this study is to deal with all these issues and steps of modeling. Thus various data mining techniques and algorithms are used to implement each step of modeling and alleviate related difficulties. For modeling purpose customers' data (30,000 customers) were gathered from Parsian bank. Based on literature and domain knowledge 85 RFM features and their two-way interactions were constructed from collected data. Since irrelevant or redundant features result in bad model performance thus feature selection was performed in order to determine the inputs to the model. Feature selection was done in three steps using F-score and backward elimination on Random Forest. The data was highly unbalanced. We used under- sampling for solving class imbalance problem. Finally SVM was used as a classifier for classification purpose. The result indicates that Parsian bank can reach three times as many respondents as if they use no model (random sampling) for target selection. By using this model Parsian bank not only can significantly reduce the overall marketing cost but also can maximize customers' response to a product offering, prevent customer annoyance and improve customer relationship management.

Place, publisher, year, edition, pages
Keyword [en]
Social Behaviour Law, Data Mining, Direct Marketing, Target Selection, Response, Model, Classification, Support Vector Machine
Keyword [sv]
Samhälls-, beteendevetenskap, juridik
URN: urn:nbn:se:ltu:diva-50999ISRN: LTU-PB-EX--08/014--SELocal ID: 837c8492-7cec-497d-9539-6719669f4c31OAI: diva2:1024362
Subject / course
Student thesis, at least 30 credits
Educational program
Electronic Commerce, master's level
Validerat; 20101217 (root)Available from: 2016-10-04 Created: 2016-10-04Bibliographically approved

Open Access in DiVA

fulltext(1463 kB)4 downloads
File information
File name FULLTEXT01.pdfFile size 1463 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search outside of DiVA

GoogleGoogle Scholar
Total: 4 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 1 hits
ReferencesLink to record
Permanent link

Direct link