![]() | Only 14 pages are availabe for public view |
Abstract Cloud computing technology has been widely spread in many fields as it has many advantages, cost saving, flexibility, quality control, and updates software automatically. Many institutions and individuals have used cloud computing technology to take benefits of cloud advantages. Individuals and organizations are storing their sensitive data, i.e., account information, financial records, and personal information in a cloud server, however, data owners do not trust the cloud server, thus data owners must be encrypted their data before outsourced to the cloud server to keep data protected from unauthorized access. Search over encrypted outsourced data can support exact search to return files that contain keyword matched with query keyword or support non exact search either fuzzy search as user may have typos like add or delete or substitutions depend on predefined threshold, or synonym search as user may use semantic keyword of stored keyword usually by using word net or semantic library, some of the search schemes support dynamic search which adds or delete file from stored index. Bloom Filter (BF) is the most commonly used in many search schemes for speed set membership tests but it does not support the update operation. In this thesis we designed a new model that uses cuckoo filter (CF) to search for an encrypted keyword stored in the cloud server. Our model improves search process by supporting update process by adding or deleting file from search index stored in the cloud server. Extensive analysis and experiments on a real-world data set showed that our model has best update time, faster in insert time and efficient in search time compared with BF (Bloom Filter). |