Search

The current web security algorithms faced some drawbacks which making them under breakdown situations, these algorithms are such as SHA-1 and MD5 which they are already hacked. This work proposes a new digital signature scheme for building a web security system that can avoid using these algorithms, so we use SHA-2 algorithm instead of them since this algorithm can be considered as not hacking algorithm at least for next years. This algorithm is currently used for database security of a client-server system, while our work exploits this algorithm for another purpose regarding to authentication process to secure users login and create new users processes. Our technology approach is proposed to use digital signature architecture based on the MRC6 algorithm to encrypt data and SHA-2 algorithm to create a digital signature by Elliptic Curve Digital Signature Algorithm (ECDSA). So, data are encrypting by the MRC6 algorithm, then extracting hash values from these data for using its inside ECDSA algorithm to generate a digital signature, after that, we merge these encrypted data with a digital signature to send them together to web server. Implemented results show that our proposed architecture can achieve the conceptual goals for web security including; confidentiality, integrity and non-repudiation, and authentication. Finally our proposed architecture is implemented using C# and SQL server.

Frequent Patterns generation takes much time to execute, compute and the efficiency is less with massive datasets. In this paper, the enhanced FP-Growth (EFP-Growth) based on MapReduce framework using Hadoop approach has been proposed. The new method has high achieving performance with different minimum support. The EFP-growth can works with the large amount of datasets to discover a frequent patterns in a transaction database. With EFP-Growth, many challenges has been solved, to mine frequent patterns such as large memory consumption, high I/O cost, high time and space complexity. In the proposed method, the execution time under different minimum supports and large datasets is decreased.

In this paper, a novelty of a new method has been presented. MRFP-Growth is implementing to solve the problem of discovering frequent patterns with massive datasets. Most of the frequent patterns algorithms are unable to manipulate with the large datasets, our scheme implement FP-Growth method under MapReduce framework in Hadoop platform to decrease the massive datasets. The efficiency and performance of a new method has been increased compare with other mining algorithms.

Prior Searchable Symmetric Encryption (SSE) works focus on single keyword search. Conjunctive Keyword Searches (CKS) schemes enhance system usability by retrieving the matching files. Most of existing conjunctive keyword works that use conjunctive keyword searches with fixed position keyword fields are not useful for many applications, such as the body of e-mail and unstructured text. The recipients have to repeatedly perform the search protocol for many times. In our paper, we propose a new symmetric key encryption scheme which supports a keyword field free method for conjunctive keyword search on encrypted file without needing to specify the positions of the keywords where the keywords can be in any arbitrary order. In this scheme, the recipients can give the server one trapdoor for multi keywords. In other words, the proposed method will retrieve the data in one round of communication between the recipients and server, and the search process could not leak any information about the number of keywords in query. Furthermore, we introduced an efficient secure index construction based on pseudorandom functions and Bloom filters. Through analysis section, we determine how such scheme could be used to guarantee fast, low storage overhead and secure access to the database.