Computer Science Homework Help

Hi, please, respond to peer discussions in 100 words minimum. The references should be in APA style and format.

Peer 1

Inference in Ordinary or Statistical Databases

A database inference attack occurs when non-sensitive public information is gathered from accessible databases and used to make inferring assumptions regarding sensitive, personal, or private data. Statistical databases are typically used to obtain new information based on previously collected or existing data sets without revealing private or individual data (Turkanović et al., 2015). Inference attacks on statistical databases, or compromises, occur when the inferences and assumptions created are tailored to specific individuals or entities (Stallings, 2017). Methods used by attackers to make inferences are typically based on “data mining, record linkage, knowledge discovery, and data analysis” (Domingo-Ferrer, 2002). Inference attacks are difficult to identify since the information used is typically general or aggregated and connections between the data must exist (Turkanović et al., 2015). Inference attacks cannot use traditional access control methods for prevention or detection since the information inferred is not unauthorized or directly accessed.

The most popular methods of protection from inference attacks are query restrictions and perturbation (Stallings, 2017). Query restrictions, or query denial, use the size of a database records to restrict query search results. For example, if a database has 10 records, the query will block access to any search results with less than 10 records. This protection puts a limitation on the ability to focus on one specific record. This protection method is useful for smaller attacks with small datasets; however, it can be weak for more complex query searches where the attacker can split his query and obtain specific characteristics from different searches. An issue with the query restriction method is that the “denial of a query may provide sufficient clues that an attacker can deduce underlying information” (Stallings, 2017). If an attacker is denied in multiple query searches, they can conclude what the restriction is and thus use it to infer information. Data perturbation produces statistical query results that cannot be used to infer specific details, while output perturbation produces modified query results that are based on original databases (Stallings, 2017). An issue with perturbation relates to the size and severity of the modification to the statistical query results provided. If the error is too small, an attacker can make inferences that are close to actual values, and if the error is too big, the query results may be invalid (Stallings, 2017).

In summary, inference attacks are a category of access, system, and computer, control that is often missed since the focus is not on direct access violations. These controls are no longer enough, and though methods of prevention and protection against inference attacks exist, they may not be as effective as required to provide continuous security of private or personal data. Inference attacks are still being researched and studied today and systems are being launched to provide increased protections and security. As inference problems continue to emerge, so will improvements and updates to the security systems that are used to protect against them.

References

Domingo-Ferrer, J. (2002). Inference Control in statistical Databases – From Theory to Practice. Springer. Retrieved from https://www.springer.com/gp/book/9783540436140.

Stallings, W. (2017, August 17). Informit. InformIT. Retrieved from https://www.informit.com/articles/article.aspx?p=782117.

Turkanović, M., Družovec, T. W., & Hölbl, M. (2015). Inference Attacks and Control on Database Structures. TEM Journal, 4(1), 3–15. Retrieved from https://www.temjournal.com/content/41/01/temjournal4101.html.

Peer 2:

Cloud Security

A relational database is a type of database management system (RDBMS) that uses related tables to store data. RDBMS’s are used commonly in modern databases because of their versatility and widespread availability. DBMS’s also have more fine grained security services and mechanisms than typical operating systems (University of Maryland Global Campus, 2021). A recent trend in the storage of large data volumes is a shift to cloud service providers. By definition, cloud computing consists of hardware, service components, networks and software which supports the development of cloud services and delivery of same via internetwork (Franklin & Ojekudo, 2021). A major concern following the shift to these cloud providers is data confidentiality.

Cloud computing confidentiality concerns include secret data loss, data leakage, and disclosing of personal data privacy (Hussein & Khalid, 2016). Solutions consist of security best practices, protecting data in transit, protecting stored data, protecting credentials, managing multiple users, and securing applications (Shazhad, 2014). These solutions are exemplified in practice by Amazon Web Services, the largest cloud service provider. The issues highlighted with these proposed solutions are mainly that cloud storage can be outsourced through a chain of providers, increasing overall vulnerability to attacks. Outsourcing can cause users to lose control of their data. Also, the transfer of cloud services between providers due to merging, government seizures, or the sale of their company can result in user files remaining inactive on multiple hard drives even after a user requests deletion of their account, increasing the timeframe in which a user’s files can be accessed and infiltrated. One future suggestion for cloud networks is for the implementation of an intrusion detection system in combination with a network intrusion prevention system, which would help to mitigate security risks relating to data leakage and attacks. This would aid in identifying, stopping and reporting abnormal traffic within the cloud storage system (Kumar, 2019).

In summary, DBMS systems make up the majority of modern databases. The responsibility of large data storage has steadily transferred to cloud service providers over the course of the last decade. The largest concern of cloud storage is data confidentiality, mainly data leakage and disclosure of personal data. Solutions for increasing confidentiality in cloud systems are security best practices, protecting data in transit, protecting stored data, protecting credentials, managing multiple users, and securing applications. Downfalls to these solutions are data outsourcing and data transfer over different providers.

References:

Franklin, M., & Ojekudo, N. A. (2021). Cloud Computing: Review of Architecture, Security Risks, Threats and Countermeasures.

Hussein, N. H., & Khalid, A. (2016). A survey of cloud computing security challenges and solutions. International Journal of Computer Science and Information Security, 14(1), 52.

Kumar, G. (2019). A review on data protection of cloud computing security, benefits, risks and suggestions. PDF). United International Journal for Research & Technology, 1(2), 26.

Shahzad, F. (2014). State-of-the-art survey on cloud computing security challenges, approaches and solutions. Procedia Computer Science, 37, 357-362. https://doi.org/10.1016/j.procs.2014.08.053