The widespread use of computing technology has been changing relationships among people in societies. Criminals are aware of the power of the technology so that many criminal activities involve more computing systems. Money laundering has been a significant criminal activity within financial computing systems for many decades. The dynamic nature of information systems has reduced the effectiveness of existing money laundering detection mechanisms that is an important challenge for societies. In this paper, we consider machine learning algorithms as complementary solutions to existing money laundering detection mechanisms. We have focused on graph-based representation of data with Node2Vec to have better classification results for money laundering detections with machine learning algorithms. Our experimental analyses show that Node2Vec enable us to select the most convenient machine learning algorithm for money laundering detections.