For second-language learners, collocational knowledge is very important. Knowing collocational phrases allows learners to speak and write in their targeted language naturally and reduce dramatically side effect of their first language. In order for learners to learn collocations easily, a lot of learning methods have been introduced. Particularly, learning from online-collocational corpus has become popular due to its accessibility and massive database. Although, its current presentation of information is still simple, it can be improved by using optimized representations in order to help users learning. In this thesis, we represent a suitable way to visualize online collocational dictionary by using graph representation in order to facilitate users’ learning and provide flexible exploration. Animation is also used to increase level of engagement for users. We use force-directed model for the layout, but we develop our own graph component and combine some current algorithms in order to create a proper algorithm for our purposes. The implementation is tested by a small group of participants and the results are promising.