Multiple Minimum Support-Based Rare Graph Pattern Mining Considering Symmetry Feature-Based Growth Technique and the Differing Importance of Graph Elements

摘要: Frequent graph pattern mining is one of the most interesting areas in data mining, and many researchers have developed a variety approaches by suggesting efficient, useful techniques integration fundamental with other advanced works. However, previous faced fatal problems that cannot consider important characteristics real world because they process both (1) different element importance (2) multiple minimum support thresholds suitable for each element. In words, elements not only frequency factors but also their own importance; addition, various composing graphs may require according to characteristics. traditional ones do such features. To overcome these issues, we propose new frequent method, which can deal thresholds. Through devised algorithm, obtain more meaningful results higher importance. We demonstrate proposed algorithm has outstanding performance compared state-of-the-art terms generation, runtime, memory usage.