One way to better understand brain function and its disorders, such as epilepsy, is to analyze the effective brain connectivity networks. A criterion is needed to identify effective brain connectivity in order to be able to estimate any effective relationships between the brain areas using the recorded functional events of the brain such as the EEG signals. After identifying the effective brain connectivity using an appropriate solution, it is possible to determine the meaningful connectivity, form an effective brain connectivity network graph, and evaluate its parameters, which are based on the graph theory. This article introduces a new way to estimate the flow of information between brain channels. For this purpose, after preprocessing the brain signals, the transfer entropy is measured between the sequence of the weighted horizontal visibility graph (WHVG) generated by the main time series, by calculating the delay parameters and then the required substitutions using the nearest neighbor error. Moreover, this method provides more valuable information for entropy compared to the horizontal visibility graph, since weighing the horizontal visibility graph makes the sudden changes more valuable that usually occur in epileptic seizures. The initial length of the intended signals is 40 seconds and the evaluation of the results based on the entropy calculation is on the sliding windows with a length of 2, 5, and 10 seconds. The entropy values are calculated on the windows and then compared with normal brain signals to better understand the results. In this article, the weighted horizontal visibility graph is compared with the horizontal visibility graph and it is found that the weighted visibility graph showed the changes better than the horizontal graph and provides more valuable information to the entropy.