Topic Maps are used for structuring contents and knowledge provided from different information sources and different languages. They are defined as semantic structures which allow organizing all the subjects they represent. They are intended to enhance navigation and improve information search in these resources. In this paper, we propose to study the quality of Topic Maps. Topic Map quality covers various aspects, some of them are common with conceptual schemas, others are common with information retrieval systems and some other aspects are specific to the problem of Topic Maps. In this paper, we have limited our work to treat the aspect of quality related to the volume of the Topic Map. In fact, Topic Maps are usually very big and voluminous, since they can contain thousands of Topics and associations. This large volume of information and complexity can lead to a bad organization of the Topic Map, so searching information using the Topic Map structure will be a very difficult task and users cannot find easily what they want. In this context, to manage the volume of the Topic Map, we propose a dynamic pruning method when we display the Topic Map by defining a list of meta-properties associated to each topic. The first meta-property represents the Topic score which reflects its relevance over the time and the second meta-property indicates the level to which belongs the Topic in the Topic Map.