International Journal of Science Engineering and Advance Technology (IJSEAT)
Not a member yet
1075 research outputs found
Sort by
A new differential private technique for frequent item mining
Get PDFFrequent itemsets mining with differential protection refers to the issue of mining all incessant itemsets whose bolsters are over a given limit in a given value-based dataset, with the imperative that the mined outcomes should not break the security of any single exchange. Current answers for this issue can't well adjust proficiency, security and information utility over vast scaled information. Toward this end, we propose a proficient, differential private incessant itemsets mining algorithm over vast scale information. In light of the thoughts of examining and exchange truncation utilizing length limitations, our algorithm decreases the algorithm force, diminishes mining affectability, and in this way improves information utility given a fixed protection spending plan
Secure File Sharing With Access Grants In Cloud
Get PDFWe first formally characterize an idea of shared ownership inside a document get to control demonstrate. We at that point propose two conceivable instantiations of our proposed shared ownership model. Our first arrangement, called Commune, depends on secure document dispersal and intrigue safe secret sharing to guarantee that all access gives in the cloud require the help of a concurred limit of owners. In that capacity, Commune can be utilized in existing mists without changes to the stages. Our second arrangement, named Comrade, influences the blockchain innovation so as to achieve accord on access control choice. In contrast to Commune, Comrade necessitates that the cloud can interpret get to control choices that achieve accord in the blockchain into capacity get to control rules, in this manner requiring minor changes to existing mists. We break down the security of our recommendations and think compare/evaluate their execution through usage using Amazon S3
A novel hybrid mechanism for credit card fraud detection on financial data
Get PDFCredit card fraud is a difficult issue in budgetary services. Billions of dollars are lost because of credit card fraud consistently. There is an absence of research thinks about on investigating genuine Master card data inferable from secrecy issues. In this project, machine learning algorithms are used to recognize credit card fraud. Standard models are right off the bat used. At that point, half and half techniques which use AdaBoost and larger part casting a voting method are connected. To assess the model viability, a freely credit card data collection is used. Then, a real-world credit card data set from a financial institution is analyzed
An Efficient and Secure Data Access Control For Cloud Storage
Get PDFWe present an official basis model of the proposed method, intended for applied cloud storage system disposition. We discourse faintness in the checking process of the session variety. Exactly, a spiteful user may modification his secret key in the meeting form, and the checking technique will flop in this case. As an extenuation, we review the key generation algorithm and improve an inspection list to perceive if the key is changed. Looking for to allay access recommendation misappropriation, we recommend CryptCloud+, are feasible consultant and revocable CPABE based cloud storage system with white-box traceability and auditing. To the top of our acquaintance, this is the originally everyday answer to protected fine-grained access control over encrypted data in cloud
A New Hybrid Method For Credit Card Fraud Detection On Financial Data
Get PDFCredit card fraud is a major issue in financial administrations. Billions of dollars are lost because of credit card misrepresentation consistently. There is an absence of research contemplates on breaking down certifiable Visa information attributable to privacy issues. In this paper, AI algorithms are utilized to identify Visa misrepresentation. Standard models are right off the bat utilized. At that point, half breed strategies which use AdaBoost and greater part casting ballot techniques are connected. To assess the model adequacy, a freely accessible credit card informational collection is utilized. At that point, a genuine Visa informational index from a money related organization is investigated. What's more, commotion is added to the information tests to further survey the robustness of the algorithms
Providing Cloud Data Security By Using Proof of Past Log
Get PDFWe propose the Cloud Log Assuring Soundness and Secrecy (CLASS) process as an elective plan for the verifying of logs in a cloud domain. In CLASS, logs are scrambled utilizing the individual client's open key so just the client can unscramble the substance. So as to anticipate unapproved alteration of the log, we create evidence of past log (PPL) utilizing Rabin's unique mark and Bloom channel. Such a methodology lessens check time fundamentally. Discoveries from our trials conveying CLASS in Open Stack exhibit the utility of CLASS in a genuine setting
A novel approach for Secure Key-Deduplication with IBBE
Get PDFWe plan a novel client-side deduplication convention named KeyD without such a free key management server by utilizing the identity-based broadcast encryption (IBBE) technique. Clients just collaborate with the cloud service provider (CSP) during the procedure of information transfer and download. Security investigation shows that KeyD guarantees information confidentiality and joined key security, and well ensures the ownership privacy simultaneously
A new analysis of distributed faulty node detection in DTNS
Get PDFPreviously proposed solutions suffer from long delays in identifying and dividing nodes producing faulty data. This is unsuitable to DTNs where nodes meet only rarely. This proposes a completely conveyed and essentially implementable way to deal with enable each DTN node to quickly distinguish whether its sensors are delivering flawed information. The dynamical conduct of the proposed algorithm is approximated by some persistent time state conditions, whose balance is portrayed. The nearness of getting out of hand nodes, attempting to bother the faulty node recognition process, is additionally considered
A New Multivariate Correlation Study for Detection of Denial-of-Service Attack
Get PDFWe present a attack detection system that utilizes Multivariate Correlation Analysis (MCA) for precise system traffic portrayal by removing the geometrical relationships between's system traffic highlights. Our MCA-based DoSattack identification framework utilizes the rule of abnormality based detection in attack acknowledgment. This makes our answer equipped for distinguishing known and obscure DoSattacks adequately by learning the examples of real system traffic as it were. Besides, a triangle-zone based system is proposed to upgrade and to accelerate the procedure of MCA. The adequacy of our proposed location framework is assessed utilizing KDD Cup 99 dataset, and the impacts of both non-standardized information and standardized information on the execution of the proposed identification framework are analyzed
A new model for product recommendation by using earlier reviews
Get PDFWe step up to the plate and concentrate the conduct attributes of early reviewers through their posted audits on two genuine huge web based business stages, i.e., Amazon and Yelp. In explicit, we isolate item lifetime into three back to back stages, to be specific early, dominant part and slouches. A client who has posted an audit in the beginning period is considered as an early analyst. We quantitatively describe early reviewers dependent on their rating practices, the accommodation scores got from others and the connection of their audits with item ubiquity. We have discovered that (1) an early analyst will in general dole out a higher normal rating score; and (2) an early commentator will in general post progressively supportive audits. Our examination of item audits additionally shows that early reviewers' evaluations and their got support scores are probably going to impact item fame. By survey audit posting process as a multiplayer rivalry diversion, we propose a novel margin-based embedding model for early analyst expectation