The calculating of Takeya similarity indexes based on the intersections and unions between two connected sets of the edge of directed graphs as the formula (1) was too complicated (The equation is abbreviated) where C(vk, vl|G) is the connected set of edge (vk, vl) in graph G. This study provided an easy calculating improved similarity indexes as the formula (2): (The equation is abbreviated) Furthermore, replacing the connected set by path set and ranking position path set, This study provided two improved similarity indexes as the formula (3) and (4): (The equation is abbreviated) (The equation is abbreviated).