Cloud Computing Patterns are Design Patterns especially for cloud applications and provide abstract solution concepts for often reoccurring problems during the implementation of cloud applications. These concepts are mainly used by developers and modelers. To learn about implemented patterns in a completed application, one has to manually analyze the code and the architecture. To improve this time-consuming method, the possibility of automating this process is investigated. This bachelor's thesis proposes an approach for a Design Pattern Detection Framework, to perform an automatic pattern detection. TOSCA, provided by OASIS, is a standardized description for the development of cloud applications. Their architectures can be described by TOSCA topologies, to model components and relationships among each other. The framework, which is developed in the context of this bachelor's thesis, is written in Java and integrated in Winery, a graphical modeling tool for TOSCA topologies, which is a part of the OpenTOSCA ecosystem. The underlying concept of this work follows an approach to detect which Cloud Computing Patterns are used in TOSCA topologies. The concept defines the modeling of Cloud Computing Patterns with TOSCA topologies and how TOSCA topologies are abstracted, to be comparable with pattern topologies. Further, the use of pattern taxonomies is explained to include the interrelations of Cloud Computing Patterns. Basically, patterns and TOSCA topologies are handled as graphs. Consequential, probabilities for possible patterns can be set. For the detection of pattern graphs in a topology graph, an algorithm for subgraph isomorphism is used.