Microservice architectures that use Message-oriented Middleware (MOM) have recently seen considerable evolution regarding extensibility, re-usability and maintainbility. Of particular interest are systems that are distributed and deployed with mixed-technologies. On one hand, such MOM-based microservice systems through their messaging middleware architecture brings flexibility to the systems. On the other hand, configuration for above systems has to quickly adapt to required changes because of continuous development process. Architecture reconstruction methods from dynamic data can keep architecture documentation and models in synchrony with the implemented architecture for such systems. However, the existing dynamic analysis methods for architecture reconstruction do not support the extraction for MOM-based microservice systems. The main challenge here is to understand and capture the asynchronous sender-receiver communication via the messaging middleware and to reconstruct the architecture model from it. In our work, we provide the ARCHI4MOM approach to automate the architecture extraction process. We instrument the sender-receiver and messaging services, collect run time data, analyse the trace data and construct the model from it. Architects can use the extracted architecture model for system refactoring and analysis of MOM-based systems. Thus, the cost and time required for manual architecture extraction process is reduced. We evaluate the accuracy of the approach by comparing the extracted model components to a manually crafted baseline model for a case study system.