New Bloom Filter Architecture For String Matching

Implementation of the Bloom filter for plagiarism detection in full text document has a problem on how to identify the same terms from different location. Location identifier can be hashed in offline mode since the collection is static. By this approach, the computation speed of the Bloom filter can be improved. Two new Bloom filter architectures are proposed in this study to overcome the problem of computational time. First architecture concatenates hash code of the string and its location identifier, while the second architecture concatenates the bit position of the string and its location identifier. Analysis was conducted to evaluate the proposed architectures in terms of computation time. From the result, computation time can be reduced if the location identifiers are hashed offline.