• Environmental Scanning in Marketing
• Machine Learning
• Modeling Competitive Interactions
• Missing Values in Market Research
• Computer-based decision support for SME’s
• Application of conjoint analysis as a marketing research tool to the development and control of future academic library services

The ability to identify and to react in time on up-and-coming chances in the business environment is an indispensable factor for the future success of an organization. Our project deals with the development of autonomous systems which support the information seeking activities of managers on a daily basis. This includes the modeling of human information seeking processes and the transfer to autonomous systems, which carry out the effective and efficient search and information retrieval of weak signals covered by various textual information in a given information environment, particularly documents available on the Internet. Our research combines approaches from various areas, such as information retrieval, data and text mining, machine learning, operations research, and management science, to support managers in their competitive and business intelligence duties and responsibilities.

Machine learning is a promising research discipline that gains an increasing importance in numerous application areas. The underlying principle of learning from examples is implemented in several classes of methods, e.g. neural networks or support vector machines, and may thus be used for knowledge discovery in marketing. Meaningful applications arise from point-of-sale scanner data analysis or customer classification in sales management for instance.

The competitor’s reactions to changing market conditions are of substantial relevance for the planning of marketing measures. Since the competitive behavior is characterized by the simultaneous use of different marketing instruments the modeling of multiple discrete choices is of particular importance. Our research activities were focused on the development of methods for analyzing and anticipating competitive reactions.

Data bases in market research, particularly those resulting from standardized surveys, are rarely complete. Therefore, the accurate handling of missing values is of substantial importance for the validity of the attained results. The main focus of our research efforts was concentrated on the development of suitable approaches for those missing values which are neither missing completely at random (“MCAR”) nor missing at random (“MAR”).

This project focused on the development of a generally accepted framework for the production of computer-based decision support systems for the marketing management of small and medium-sized enterprises. Following a transdisciplinary approach we were bundling current case studies and research results of different knowledge domains in a goal-oriented way.

Please see Completed Projects