Domain integrated product recommendations for a B2B enterprise: A Unified Recommender System

back to overview

Type and Duration

Preproposal PhD-Thesis, since September 2021

Coordinator

Hilti Chair for Data and Application Security

Main Research

Business Process Management

Description

Significant changes in the consumer's purchasing behavior have revolutionized the development of recommender systems. B2C enterprises have been at the forefront in producing the most advanced algorithms and utilizing recommender systems to improve customer experience. On the other hand, the complex hierarchical structure of the consumer, multi-domain business models, and increasing communication channels have hindered B2B enterprises in capitalizing on these systems. The focus of the research will be
to use Machine Learning models to generate recommendation and the best practices to disseminate these recommendations to the customers. The document proposes a cumulative and paper-based dissertation project, "Domain Integrated Product Recommendations for a B2B Enterprise: A Unified Recommender System", with the Hilti
Chair of Data and Application Security at the University of Liechtenstein. The following sections of the proposal describe the need for the research, existing literature, core problem
statements, and quantitative research methodology to be used during the research course.
Digitalization over the past decade has dramatically changed the purchasing behavior of consumers. Amazon, selling just books in 1995 to racing towards a $2 trillion giant in 2021 (Jeff Bezos and Amazon.Com by Mark E. Parry :: SSRN), narrates the growth of e-commerce in the internet age. The last few months with the covid-19 pandemic have accelerated the ecommerce industry by 4-6 years, increasing the sales by 77% compared to the sales of the
past year(Nanasaheb Tayade, Khandeshwar and Amravati, 2021).

The growth in the e-commerce industry has fuelled competition, making more companies become part of the global market. An increase in choices for a consumer has become a challenge for enterprises to retain customers, and hence, in this context, a recommender system comes across as one of the efficient solutions.

Business-to-Consumer (B2C) companies have effectively developed and used recommender systems. From a novelty at an e-commerce site, these recommendation algorithms - suggesting products and services - have become serious business tools(Schafer, Konstan and
Riedl, 1999). Business to Business (B2B) companies have tried to capitalize on the same technology, but the intrinsic behavior of a business as a consumer has not made it easy.
Utilizing recommender systems for B2B companies with multiple business solutions and increasing communication channels is extremely difficult, making it a focus area of the dissertation.

Digitalization provides a challenging but unique opportunity to B2B companies operating with a direct sales model. The convergence of a massive physical sales infrastructure (people and stores) with an understanding of customer's digital footprint can generate better sales opportunities. A coherent consolidation of physical and digital will serve as the secondary area of research for the dissertation.