Similar files result strong different

[alee001]

Every time we use the similarity graph to compare, not only the CPU resource usage is high, but the results are also very different. (Example in the pictures)
Hope the next version will improve the algorithm...