Customers interested in buying a product, can search on the internet for reviews about that product. For many products, an enormous amount of information and opinions is available. Customers gets overwhelmed by this information and systems are needed to filter out the essential information. In this research, a model is developed to automatically create a graphical summary of a set of Dutch product reviews. This model is divided into three sub modules. The first module is meant to select sentences expressing opinions. The second module groups those sentences based on their subjects. Finally, the third module generates a graphical representation of the groups of sentences and presents it to the customer. The most important parts of the model are implemented into a system. An user experiment is conducted to evaluate the performance of this system. This experiment shows that the first module has an accuracy of 81%. This is comparable with systems developed for other languages. Information shown in the graphical summary representation is judged as relevant in 56% of the cases. Although further research is needed, this model is a good basis to provide insight to the customer about essential information in product reviews.