Model-based Purchase Predictions for Large Assortments

44 Pages Posted: 29 May 2014

See all articles by Bruno Jacobs

Bruno Jacobs

Erasmus University Rotterdam (EUR) - Erasmus School of Economics (ESE); Erasmus Research Institute of Management (ERIM)

Bas Donkers

Erasmus University Rotterdam (EUR) - Erasmus School of Economics (ESE); Erasmus Research Institute of Management (ERIM); Tinbergen Institute

D. Fok

Econometric Institute - Erasmus University Rotterdam; Erasmus Research Institute of Management (ERIM); Tinbergen Institute Rotterdam

Date Written: February 18, 2016

Abstract

Being able to accurately predict what a customer will purchase next is of paramount importance to successful online retailing. In practice, customer purchase history data is readily available to make such predictions, sometimes complemented with customer characteristics. Given the large assortments maintained by online retailers, scalability of the prediction method is just as important as its accuracy. We study two classes of models that use such data to predict what a customer will buy next: A novel approach that uses latent Dirichlet allocation (LDA), and mixtures of Dirichlet-Multinomials (MDM). A key benefit of a model-based approach is the potential to accommodate observed customer heterogeneity through the inclusion of predictor variables. We show that LDA can be extended in this direction while retaining its scalability. We apply the models to purchase data from an online retailer and contrast their predictive performance with that of a collaborative filter and a discrete choice model. Both LDA and MDM outperform the other methods. Moreover, LDA attains performance similar to that of MDM while being far more scalable, rendering it a promising approach to purchase prediction in large assortments.

Keywords: model-based recommendations, product recommendations, scalability, topic models, latent Dirichlet allocation

Suggested Citation

Jacobs, Bruno and Donkers, Bas and Fok, Dennis, Model-based Purchase Predictions for Large Assortments (February 18, 2016). ERIM Report Series Reference No. ERS-2014-007 MKT. Available at SSRN: https://ssrn.com/abstract=2443455 or http://dx.doi.org/10.2139/ssrn.2443455

Bruno Jacobs (Contact Author)

Erasmus University Rotterdam (EUR) - Erasmus School of Economics (ESE) ( email )

P.O. Box 1738
3000 DR Rotterdam, NL 3062 PA
Netherlands

Erasmus Research Institute of Management (ERIM) ( email )

P.O. Box 1738
3000 DR Rotterdam
Netherlands

Bas Donkers

Erasmus University Rotterdam (EUR) - Erasmus School of Economics (ESE) ( email )

P.O. Box 1738
Rotterdam, NL 3000 DR
Netherlands
+31 10 408 2411 (Phone)
+31 10 408 9169 (Fax)

HOME PAGE: http://people.few.eur.nl/donkers/

Erasmus Research Institute of Management (ERIM) ( email )

P.O. Box 1738
3000 DR Rotterdam
Netherlands
+31 10 408 2411 (Phone)
+31 10 408 9169 (Fax)

HOME PAGE: http://people.few.eur.nl/donkers/

Tinbergen Institute ( email )

Burg. Oudlaan 50
Rotterdam, 3062 PA
Netherlands

Dennis Fok

Econometric Institute - Erasmus University Rotterdam ( email )

P.O. Box 1738
3000 DR Rotterdam
Netherlands

Erasmus Research Institute of Management (ERIM) ( email )

P.O. Box 1738
3000 DR Rotterdam
Netherlands
+31 10 408 1333 (Phone)
+31 10 408 9162 (Fax)

Tinbergen Institute Rotterdam ( email )

P.O. Box 1738
3000 DR Rotterdam
Netherlands

Register to save articles to
your library

Register

Paper statistics

Downloads
251
Abstract Views
1,258
rank
121,777
PlumX Metrics