Abstract
Research into service provision and innovation is becoming progressively more important as automated service-provision via the web matures as a technology. We describe a web-based targeting platform that uses advanced dynamic model building techniques to conduct intelligent reporting and modeling. The impact of the automated targeting services is realized through a knowledge base that drives the development of predictive model(s). The knowledge base is comprised of a rules engine that guides and evaluates the development of an automated model-building process. The template defines the model classifier (e.g., logistic regression, multinomial logit, ordinary least squares, etc.) in concert with rules for data filling and transformations. Additionally, the template also defines which variables to test (“include” rules) and which variables to retain (“keep” rules). The “final” model emerges from the iterative steps undertaken by the rules engine, and is utilized to target, or rank, the best prospects. This automated modeling approach is designed to cost-effectively assist businesses in their targeting activities—independent of the firm’s size and targeting needs. We describe how the service has been utilized to provide “targeting services” for a small to medium business direct marketing campaign, and for direct sales-force targeting in a larger firm. Empirical results suggest that the automated modeling approach provides superior “service” in terms of cost and timing compared to more traditional manual service provision.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
For SMB file—uploads and CRM “on-the-fly” targeting, Fig. 2 displays the basic workflow. For larger enterprise clients, the customer data would already be loaded (first two steps would be complete). In the enterprise case, the user would create a customer segment of interest (rather than upload it); otherwise, the process would be displayed as in Fig. 2.
In the case of an enterprise installation, the automated process would ease subsequent segment creation for subsequent modeling projects. For the initial model, however, there is little difference between the automated and manual processes.
Of course, the deep structure embedded in the modeling templates requires this same expert statistical training, but the automated system allows “re-use” of this training by other users.
Purchase terms may differ but typically for manual models the entire set of 0’s (along with demographic data) would need to be purchased before modeling while for the automated process only the actionable ranked list is purchased for deployment.
Kridel and Dolk (2003) provides examples of system artifacts and a simplified version of a modeling template (in flow-chart form).
Note these include rules can apply to individual categories (within the variable) or to all categories (within the variable).
These are slightly more general in the sense that the rules can be applied for the entire variable (or by components of the variables). In the case of categorical variables, the rule can be applied to all categories (keep all if any are significant) or by category (keep only the categories that satisfy the keep rule). For continuous variables, the same options can be applied (keep all functional terms if any pass keep rules or keep only the functional terms that pass keep rules). In addition, the keep rules can be applied at each step or only in the current step.
MOSAIC is a geo-demographic segmentation system developed by Experian. (There are several alternative segmentation systems available from other vendors.) Each of the nearly 250,000 block groups in the US are categorized into one of 12 groups which can be further broken down into 60 segments (MOSAIC codes), e.g., A01 = America’s Wealthiest, A02 = Dream Weavers, etc.
When there are continuous variables, the process also tests for functional form; that is, does the variable enter as linear, quadratic, logarithmic, etc. The test may be based on categorical equivalents (the “shape” of the univariate report) or likelihood-based tests based on Box-Cox transformations.
This assumes the model test was satisfied; if not, the final model would be altered based on the model rules in place and parts of the iterative process would be repeated.
The targeting platform was initially called Bizfusion, a proprietary product of CopperKey, Inc. In 2009, KAST acquired the IP and renamed the service. Since the original submission of this article, KAST has been acquired by Adenyo, Inc. and the platform will likely be renamed once again.
It is worth noting that the modeling profile was used for the “dumb or intuitive” selects. As a result, the “intuitive” list is better than it would have been had the business owner simply had to guess at the appropriate selects.
References
Bhargava H, Krishnan R, Muller R (1997) Decision support on demand: emerging electronic markets for decision technologies. Decis Support Syst 19:193–214
Castillo D, Dolk D, Kridel D (1991–1992) GOST: an active modeling system for costing and planning NASA space programs. J Manag Inf Syst 8(3):151–169
Chomsky N (1957) Syntactic structures. Mouton Press
Dietrich B (2006) Resource planning for business services. Commun ACM 49(7):62–64
Dolk D (2008) Introduction to decision technologies and service sciences track. In: Proceedings of the 41st Hawaiian international conference on system sciences, January 2008. IEEE Computer Society
Dolk D, Kridel D (1991) An active decision support system for econometrics. Decis Support Syst 7:315–328
Dolk D, Kridel D (1993) Towards a symbiotic expert system for econometric modeling, Chap. 7. In: Blanning RW, King DR (eds) Current research in decision support technology. IEEE Computer Society Press
Edmiston E, Kridel D (2007) Automated modeling and sales targeting: case study for AT&T advertising and publishing. In: Proceedings of DMA07 (Direct Marketing Association conference and exhibit), Chicago, IL, October 2007
El-Gayar O, Deokar A (2011) An ontology-based model management architecture for service innovation. In: Dolk D, Granat J (eds) Modelling for decision support in network-based services. Lecture Notes in Business Information Processing. Springer, Berlin (to appear)
Feng L, Dillon T (2005) An XML-enabled data mining query language: XML-DMQL. Int J Bus Intell Data Min 1(1):22–41
Geoffrion A, Maturana S (1995) Generating optimization-based decision support systems. In: Proceedings of the 28th Hawaii international conference on system sciences. IEEE Computer Society
Guazzelli A, Lin W, Jena T (2010) PMML in action: unleashing the power of open standards for data mining and predictive analytics. CreateSpace
Han J, Fu Y, Koperski K, Wang W, Zaiane O (1996) DMQL: a data mining query language for relational databases. In: Proceedings of the ACM SIGMOD workshop on research issues on data mining and knowledge discovery, June 1996
Kannan P, Proenca J (2008) Design of service systems under variability: research issues. In: Proceedings of the 41st Hawaiian international conference on system sciences, January 2008. IEEE Computer Society
Kridel D, Dolk D (2003) An on-line marketing consultant for small and medium businesses. In: Proceedings of the 32nd WDSI conference, Lihue, HI, April 2003
Kridel D, Dolk D (2004) Using intelligent profiling to generate smart lists: an empirical test. In: Proceedings of the 33rd WDSI conference, Manzanilla, Mexico, April 2004
Kridel D, Dolk D (2009) A self-service automated targeting portal: an example of model as a service. In: Proceedings of the 38th WDSI conference, Lihue, HI, April 2009
Maglio P, Srinivasan S, Kreulen J, Spohrer J (2006) Service systems, service scientists, SSME, and innovation. Commun ACM 49(7):81–85
Manheim M (1988) An architecture for active DSS. In: Proceedings of the 21st Hawaiian international conference on system sciences, vol III, January 1988. IEEE Computer Society, pp 356–365
Muthukumar A, Nadarajan R (2007) TDML: a data mining language for transaction databases. Fourth international conference on fuzzy systems and knowledge discovery, pp 81–86
Rai A, Sambamurthy V (2006) Editorial notes—the growth of interest in services management: opportunities for information systems scholars. Inf Syst Res 17(4):327–331
Rossi P, McCulloch R, Allenby G (1996) The value of purchase history data in target marketing. Mark Sci 15(4):321–340
Rust R, Miu C (2006) What academic research tells us about service. Commun ACM 49(7):49–54
Shaffer G, Zhang Z (1995) Competitive coupon targeting. Mark Sci 14(4):395–416
Spohrer J, Riiecken D (2006) Services science. Commun ACM 49(7):31–34
Spohrer J, Maglio P, Bailey J, Gruhl D (2007) Towards a science of service systems. Computer 40(1):71–77
Susarla A, Barua A, Whinston AB (2003) Understanding the service component of application service provision: an empirical analysis of satisfaction with asp services. MIS Q 27(1):91–123
Two Crows Corporation (2005) Introduction to data mining and knowledge discovery, 3rd edn. http://www.twocrows.com/intro-dm.pdf
Vargo SL, Lusch RF (2004) Evolving to a new dominant logic for marketing. J Mark 68(1):1–17
Wand Y, Weber R (1995) On the deep structure of information systems. Inf Syst J 5:203–223
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kridel, D., Dolk, D. Automated self-service modeling: predictive analytics as a service. Inf Syst E-Bus Manage 11, 119–140 (2013). https://doi.org/10.1007/s10257-011-0185-1
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10257-011-0185-1