Residue Hash法

概要

論文の詳細を見る
In this paper, a new hashing technique for multicomponent data is proposed. When data have n components, they should be stored in a n-dimensional array so that they may be easy to retrieve using less than n keys. Ordinarry arrays used in FORTRAN or ALGOL are not suitable for this purpose, because they are not symmetrial among their indexes. In the residue hash method, introduced in the present paper, this array corresponds with residue number representation of the data in n-base residu system, that is each component of data is divided by corresponding base and the remainder is used for the index of the array. This type of array is symmetrical among its indexes. Conflitcs in hashing are easily processed by linear conflict processing using the following modulo equation. X+1=(x_1+1,x_2+1,.....,+x_n+1) mod Π__ib_i, where (x_1,x_2,......,x_n) is a residue representation of X and b_i's are bases of the residue system. Efficiency in storing and searching data by this method is discussed.
一般社団法人情報処理学会の論文
1971-03-15