In the NEXT project, those methods and approaches are researched on an application-oriented level that can be used to obtain data literacy in the retail sector of the future. As part of the research project, research is carried out on an operational and tactical level using methods of supervised learning (e.g. support vector machines) and unsupersvised learning (e.g. neural networks) to find out which patterns can be recognized in data of complex value-added networks and how these can be used as the basis for predictive statements and recommendations for action can be prepared and implemented. On the other hand, using data and web mining at a strategic level, research is carried out into the form in which signals from which relevant (new) input sources (qualitative, quantitative) are collected and analyzed to reduce uncertainties, can be processed and used as part of long-term network planning. In the NEXT project, there are two focal target areas: the area "pattern recognition" and the area "data or web mining":

• In the target area “pattern recognition”, machine learning algorithms are applied to large amounts of data from completed transactions in the value-added network. In depth analysis of this historical data from defined areas of the value-added network enables patterns to be recognized and early warning indicators to be derived. The aim is to be able to make predictive statements about threatening, critical events in the preliminary stages of possible deviations between the TARGET and ACTUAL state of the value-added network based on "past learnings".
• In the target area "Data or Web Mining", information and knowledge gaps are identified in defined tactical and strategic SCM processes and analyzed for their relevance. By combining data mining techniques and machine learning approaches, critical information gaps should be closed and medium and long-term SCM should be able to build on more comprehensive and well-founded databases.