This study introduces a novel authentication methodology; it is based on pattern recognition of fingers size and pressure when users touch smartphone screen. By analyzing diagrams of these touches and applying data mining for the first time as an authentication technique, this paper presents three new approaches. First, an exact-range evaluation approach has been verified that size is more recognition consistency than pressure. Second, a pattern-range is a new technique reliance on size frequency position. At last, using a size-range has been facilitated the login. The association rules have been modified to work on finger touchscreen data files. To login, 94.1111% of 18 authorized users are succeeded and 98.9% of 20 unauthorized users are failed. Android device and Android studio are used. Size and pressure are normalized to 1; a training set is applied; the password is not considered.