Abstract: Semantic wikis have been regarded as an important collaboration tool among a number of experts from multiple domains. This wiki platform can play a role of collaborative knowledge management system which can provide an efficient framework to raise social interactions between remote people synchronously. However, as these semantic wiki systems allow users to exploit their own semantics and backgrounds for describing their knowledge and skills, there are often semantic conflicts between knowledge (or information) published and provided by the users. Thereby, the main aims of this work are i) to automatically detect such conflicts by keeping track on the user semantics, and ii) to reasonably select consensus choice by analyzing social collaborations. In this paper, we want to note major patterns of knowledge dynamics through the social interactions on semantic wikis, and the semantic conflicts caused by the knowledge dynamics. The consensus choice has been effectively selected to be recommended for better understandability about the knowledge conflicts.