Processes in manufacturing and logistics are characterized by a high frequency of changes and fluctuations, caused by the high number of participants in logistic processes and the variety of goods handled and services offered. This dynamic behavior particularly requires well documented processes, but at the same time it also complicates process documentation. Manually documented processes can, e.g., miss details of alternative process branches, and the continuing change in logistic processes renders the documentation quickly outdated. A possible solution approach is to automate the documentation of processes. This automated documentation, if based on transaction and master data from the IT systems connected to the logistic process, is called Process Mining. Being a subfield of Data Mining, Process Mining extracts sequences of activities from event logs in databases. Logistics has the opportunity to greatly benefit from the application of Process Mining, because the identification and tracking of goods in the supply chain involves many IT systems. However, the IT landscape in logistics is heterogeneous, because the data are scattered among different specialized systems for various purposes (e.g., warehousing, transportation planning, and billing) of different companies. Due to this lack of standardization, the data cannot simply be analyzed by a predefined routine. Therefore, additional information beside the stored data should be taken into consideration. Machine Learning offers the possibility to classify single items in large data sets and to categorize these items with regard to the context they are in. Adding context awareness to unstructured event data in logistics has the potential to improve the results of Process Mining. Our research investigates how to apply context awareness based on Machine Learning in Process Mining for logistic processes and demonstrates its performance in a logistic scenario.