Gerold Kreutler

Personalized Needs Elicitation in Web-based Configuration Systems

The high product variety of a mass customization strategy induces a high level of complexity both from the mass-customizer’s perspective as well as from the customers’ viewpoint. In particular, a high number of different product variants and configurable features can be challenging for the end-user who is often overwhelmed during the configuration and buying process. As customers are generally not technical engineers, but rather less-experienced, they are often confused and unable to choose the product that best fits their needs. As a consequence, customers can be dissatisfied with their buying decision later on, which finally leads to frustration and a decrease of customer loyalty. Web-based product configuration systems are nowadays well-established in commercial environments and enable users to specify desired product variants typically on a technical level. Thus, they efficiently support product experts in configuring their desired product variant. However, most current systems do not take into account the fact that online configuration systems should be usable and helpful for quite heterogeneous user groups. Online customers typically have a different background in terms of experience or skills or are simply different in the way they prefer to (are able to) express their needs and requirements. Thus, we argue that the typical “one-style-fits-all” approach for needs elicitation is not adequate for customer-supplier-interaction in mass customization. As users are different, it is necessary to adapt the interaction to the customer, i.e. to take the user’s background or his capabilities into account and tailor the interaction accordingly. Within this paper, we comprehensively discuss personalization and adaptation possibilities for interactive needs elicitation in online configuration by categorizing the different levels and dimensions in a conceptual framework. Throughout, we describe adequate techniques for effectively implementing such functionality and give examples for personalization opportunities for the different levels. Finally, we discuss architectural aspects when building and maintaining such highly-adaptive web applications. Our work extends already existing work on personalization for product configuration systems. However, while most existing approaches base their adaptation features on long-term user models, we focus on (knowledge-based) techniques that allow us to personalize the interaction style also for first-time users, for which there is nearly no support in most existing systems.

A Knowledge-Based Framework for the Rapid Development of Conversational Recommenders

Web-based sales assistance systems are a valuable means to guide online customers in the decision-making and product selection process. Conversational recommenders simulate the behavior of an experienced sales expert, which is a knowledge-intensive task and requires personalized user interaction according to the customers’ needs and skills. In this paper, we present the Advisor Suite framework for rapid development of conversational recommenders for arbitrary domains. In the system, both the recommendation logic and the knowledge required for constructing the personalized dialog and adaptive web pages is contained in a declarative knowledge-base. The advisory application can be completely modeled using graphical tools based on a conceptual model of online sales dialogs. A template mechanism supports the automatic construction of maintainable dynamic web pages. At run-time, a controller component generically steers the interaction flow. Practical experiences from several commercial installations of the system show that development times and costs for online sales advisory systems can be significantly reduced when following the described knowledge-based approach.

ADVISOR SUITE: A Tool for Rapid Development of Maintainable Online Sales Advisory Systems

A sales advisory system is a tool supporting customers in the decision-making and buying process by interactive and personalized requirements elicitation and the provision of comprehensible product proposals and explanations. The particular challenges when building such systems lie in the strong interdependencies between the recommendation and personalization logic and the corresponding adaptive, web-based user interface. The Advisor Suite tool described in this paper is a system that follows a consistent knowledge-based approach for all tasks that are required to build such intelligent sales advisory systems for arbitrary domains. The development of advisory applications is based on a conceptual model of online sales dialogues, a “Model-View-Controller” application architecture, a generic controller component, as well as (semi-)automatic, template-based web page generation. Experiences from various real-world applications show that the knowledge-based approach and the corresponding graphical tools of Advisor Suite significantly accelerate the development and maintenance process for such applications.