Developing a CBR System for Marketing Mix Planning and Weighting Method Selection Using Fuzzy AHP

Figures & data

TABLE 1 CBR Studies in Marketing and Sales Domain

Study	Application Domain
McSherry (2005)	Product recommendation
Sun and Finnie (2005)	E-commerce
Chang and Lai (2005)	Sales forecasting
Changchien and Lin (2005)	Marketing planning
Dong (2006)	Customer segmentation
Juan et al. (2006)	Pre-sale housing
Ahn, Kim, and Han (2007)	Customer classification
Lee et al. (2007)	Recommendation system with marketing orientation
Chang et al. (2008)	Sales forecasting
Kim and Ahn (2008)	Market segmentation
Althuizen and Wierenga (2008)	Design of creative sales promotion campaigns
Heras et al. (2009)	Solution recommender for customer support
Juan et al. (2010)	Pre-sale housing
Chen, Wang, and Feng (2010)	Market segmentation and developing marketing strategies
Ahn et al. (2011)	Cross-selling in a mobile telecom
Büyüközkan and Ergün (2011)	Recommendation tool for trip planning (electronic tourism)
Noori (2012)	Proposal generation for sports marketing
Lee et al. (2012)	Sales assessment
Vukovic et al. (2012)	Credit scoring and classification
Liu and Wang (2012)	Sales forecasting

TABLE 2 Input Features of the CBR System

Feature	Data Type	Content
Industrial sector	Character	Metal; Automotive; Home appliance; Electrical; Machinery; …
Product application	Character	Cutting; Welding; Painting; Cutting and bending; Stamping; …
Image and liability	List	Profitability
Credit	List	Long-term liability
Order quantity	List	Purchase amount
Strength points	Character	S1: Fixed assets turnover; S2: Number of employees
Weakness points	Character	W1: Geographical position; …

TABLE 3 Output Features of the CBR System

Solution	Data Type	Content
Product	Memo	Thickness; Standard; Quality
Promotion strategy	Text	Extended warranty; Freightage waiver; Price-pack deal; …
Relationship policy	Integer	0: Short term; 1:Long term
Pricing strategy	Text	Skimming; Loss leader; Freemium; Decoy pricing; Marginal-cost pricing; Market-oriented pricing; Penetration pricing; …
Meeting and proposal points	Memo	Proposals; Contracts; Meeting notes
Packaging	Text	Details
Delivery	Text	Descriptions

TABLE 4 Sample Dataset

Attributes	Company A	Company B	Company C	Company D
Features
Sector	Automobile parts	Metal products	Home Appliances	Buildings
Product applications	Stamping	Manufacture of tanks (Cutting and welding)	Bending (Hoods)	Painting (Industrial buildings)
Image and liabilities	Normal	Weak	Normal	Good
Credit	Good	Good	Good	Good
Order Quantity	High	Medium	High	High
Strength Points	S1	−	S1	S1,S2
Weakness Points	−	−	W2	−
Solutions/Advises
Product	Commercial steel sheet	A526, A653-CQ	A366-91, A619, A620 2.00˜3.00mm	A446-A, B, D, E, F, A653-SQ230, SQ255
Promotion strategy	Gifts and Samples	Tie-in	Extended warranty	Freightage waiver
Relationship policy	Medium- term	Short-term	Long-term	Long-term
Pricing strategy	Cost plus	Skimming	Loss leader	Penetration
Meeting and proposal points	Contract-C120	Agenda C154	Proposal-PC54	Agenda C230
Packaging	Normal package	Normal package	Standard export package, in 20 feet container	Standard export package, in 20 feet container
Delivery	About 20 days	About 20 days	About 10 days	About 15 days

FIGURE 1 CBR system architecture.

FIGURE 2 CBR dynamic weight vector assignment, rectangular dashed line.

FIGURE 3 Retrieval user interface screen.

TABLE 5 Illustration of the Prediction Evaluation Function (Y)

Training Cases	Expected Outcome (O^′j)	Real Outcome (Oj)	Matched (yj)
Case1	Yes	Yes	1
Case2	No	Yes	0
⋮	⋮	⋮	⋮
Casej	No	No	1
⋮	⋮	⋮	⋮
Casep	⋮	⋮	⋮

TABLE 6 The GA Parameters

Population size	30
Crossover rate	0.8
Mutation rate	0.3
Criteria for stopping process/Number of loops	40

TABLE 7 Output of GA System under Different Parameter Sets

Iteration	W1	W2	W3	W4	W5	W6	W7	Weight Efficiency
Set 1	0.097663	0.012266	0.197414	0.212537	0.163813	0.081075	0.036598	74.286
Set 2	0.154967	0.037759	0.035773	0.081574	0.173436	0.167232	0.133933	78.441
Set 3	0.156657	0.152547	0.085575	0.051429	0.023571	0.143704	0.242292	77.143
Set 4	0.006844	0.168711	0.18918	0.11584	0.111154	0.032386	0.017021	77.143
Set 5	0.031975	0.022602	0.060525	0.190025	0.181494	0.116027	0.01376	77.143
Set 6	0.184884	0.076168	0.044178	0.009155	0.107875	0.237772	0.081619	77.143
Set 7	0.270192	0.133757	0.110788	0.093291	0.081051	0.044384	0.088663	77.143
Set 8	0.04642	0.09305	0.232842	0.140804	0.160858	0.093848	0.043171	74.286
Set 9	0.078648	0.047989	0.048015	0.102393	0.268818	0.189285	0.350423	77.143
Set 10	0.18437	0.159735	0.084425	0.033344	0.052807	0.106705	0.095163	71.429

TABLE 8 Output of the GACBR System

Weight	Values
W1	0.154967
W2	0.037759
W3	0.035773
W4	0.081574
W5	0.173436
W6	0.167232
W7	0.133933

TABLE 9 Output of the RSCBR System

Weight	Values
W1	0.176
W2	0.235
W3	0.117
W4	0.117
W5	0.117
W6	0.117
W7	0.117

FIGURE 4 Weight vector based on RSCBR, rectangular dashed line.

FIGURE 5 Partition of monetary universe set.

FIGURE 6 Partition of frequency universe set.

FIGURE 7 Partition of output universe set.

FIGURE 8 Mamdani FIS application scheme, followed in this study.

FIGURE 9 The flowchart of FIS methodology.

FIGURE 10 Fuzzy reasoning for multiple rules with antecedents.

TABLE 10 Sample Output of FIS

Weight	Values
W1	0.05
W2	0.05
W3	0.18
W4	0.26
W5	0.16
W6	0.12
W7	0.18

FIGURE 11 Problem hierarchy.

TABLE 11 The Linguistic Variables and Their Corresponding Fuzzy Numbers

Equally preferred (EP)	(1, 1, 1)
Weakly preferred (WP)	(2/3, 1, 3/2)
Fairly strongly preferred (FSP)	(3/2, 2, 5/2)
Very strongly preferred (VSP)	(5/2, 3, 7/2)
Absolutely preferred (AP)	(7/2, 4, 9/2)

TABLE 12 The Fuzzy Evaluation Matrix with Respect to the Goal with Linguistic Variables

	Accuracy	Adoption	Implementation	Compatibility
Accuracy			FSP
Adoption	FSP		FSP	FSP
Implementation				WP
Compatibility	WP

TABLE 13 The Fuzzy Evaluation Matrix

	Accuracy	Adoption	Implementation	Compatibility
Accuracy	(1, 1, 1)	(2/3, 1, 3/2)	(3/2, 2, 5/2)	(2/3, 1, 3/2)
Adoption	(2/3, 1, 3/2)	(1, 1, 1)	(2/3, 1, 3/2)	(3/2, 2, 5/2)
Implementation	(2/5, 1/2, 2/3)	(2/3, 1, 3/2)	(1, 1, 1)	(2/3, 1, 3/2)
Compatibility	(2/3, 1, 3/2)	(2/5, 1/2, 2/3)	(2/3, 1, 3/2)	(1, 1, 1)

TABLE 14 Priority Vectors for the Decision Hierarchy

Variables in Level 1	Level 1 Priorities	Variables in Level 2	Level 2 Priorities
Accuracy	0.33	GA	0.32
		RST	0.31
		Fuzzy System	0.37
Adoption	0.45	GA	0.22
		RST	0.33
		Fuzzy System	0.45
Implementation	0.02	GA	0.33
		RST	0.22
		Fuzzy System	0.45
Compatibility	0.19	GA	0.16
		RST	0.84
		Fuzzy System	0

TABLE 15 Weighting System Priorities

	Accuracy	Adoption	Implement-ation	Compatibility	Alternative Priority Weight
Weighting	0.33	0.45	0.02	0.19
Alternatives
GA	0.32	0.22	0.33	0.16	0.24
RST	0.31	0.33	0.22	0.84	0.41
Fuzzy System	0.37	0.45	0.45	0	0.33

McSherry, D. 2005. Explanation in recommender systems. Artificial Intelligence Review 24(2):179–197.

 Web of Science ®Google Scholar

Sun, Z., and G. Finnie. 2005. A unified logical model for CBR based e-commerce systems. International Journal of Intelligent Systems 20(1):29–46.

 Web of Science ®Google Scholar

Chang, P.-C., and C.-Y. Lai. 2005. A hybrid system combining self-organizing maps with case-based reasoning in wholesaler’s new-release book forecasting. Expert Systems with Applications 29(1):183–192.

 Web of Science ®Google Scholar

Changchien, S., and M. C. Lin. 2005. Design and implementation of a case-based reasoning system for marketing plans. Expert Systems with Applications 28(1):43–53.

 Web of Science ®Google Scholar

Dong, Y. 2006. A case based reasoning system for evaluating customer credit. Journal of Japan Industrial Management Association 57(2):144–152.

 Google Scholar

Ahn, H., K. Kim, and I. Han. 2007. A case-based reasoning system with the two-dimensional reduction technique for customer classification. Expert Systems with Applications 32(4):1011–1019.

 Web of Science ®Google Scholar

Lee, H.-Y., H. Ahn, and I. Han. 2007. VCR: Virtual community recommender using the technology acceptance model and the user’s needs type. Expert Systems with Applications 33(4):984–995.

 Web of Science ®Google Scholar

Chang, P.-C., C.-H. Liu, and R.K. Lai. 2008. A fuzzy case-based reasoning model for sales forecasting in print circuit board industries. Expert Systems with Applications 3(3):2049–2058.

 Google Scholar

Ahn, H., and K. Kim. 2008. Using genetic algorithms to optimize nearest neighbors for data mining. Annals of Operations Research 163(1):5–18.

 Web of Science ®Google Scholar

Althuizen, N. A. P., and B. Wierenga. 2008. The value of analogical reasoning for the design of creative sales promotion campaigns: A case-based reasoning approach. Working paper.

 Google Scholar

Heras, S., J. Á. García-Pardo, R. Ramos-Garijo, A. Palomares, V. Botti, M. Rebollo, and V. Julián. 2009. Multi-domain case-based module for customer support. Expert Systems with Applications 36(3):6866–6873.

 Web of Science ®Google Scholar

Juan, Y.-K., S.-F. Chien, and Y.-J. Li. 2010. Customer focused system for pre-sale housing customisation using case-based reasoning and Feng Shui theory. Indoor and Built Environment 19(4):453–464.

 Web of Science ®Google Scholar

Chen, Y. K., C.Y. Wang, and Y. Y. Feng. 2010. Application of a 3NN+ 1 based CBR system to segmentation of the notebook computers market. Expert Systems with Applications 37(1):276–281.

 Web of Science ®Google Scholar

Ahn, H., J. J. Ahn, K. J. Oh, and D. H. Kim. 2011. Facilitating cross-selling in a mobile telecom market to develop customer classification model based on hybrid data mining techniques. Expert Systems with Applications 38(5):5005–5012.

 Web of Science ®Google Scholar

Büyüközkan, G., and B. Ergün. 2011. Intelligent system applications in electronic tourism. Expert Systems with Applications 38(6):6586–6598.

 Web of Science ®Google Scholar

Noori, B. 2012. Marketing DSS architecture for sports sponsorship management. Journal of Database Marketing and Customer Strategy Management 19(4):29–299.

 Google Scholar

Lee, N., J.K. Bae, and C. Koo. 2012. A case-based reasoning based multi-agent cognitive map inference mechanism: An application to sales opportunity assessment. Information Systems Frontiers 14(3):653–668.

 Web of Science ®Google Scholar

Vukovic, S., B. Delibasic, A. Uzelac, and M. Suknovic. 2012. A case-based reasoning model that uses preference theory functions for credit scoring. Expert Systems with Applications 39(9):8389–8395.

 Web of Science ®Google Scholar

Liu, C., and H. Chen. 2012. A novel CBR system for numeric prediction. Information Sciences 185(1):178–190.

 Web of Science ®Google Scholar