A presented a revolutionary system called Spark that makes use of Apache Spark to efficiently manage large amounts of health care data as well as K-anonymization and L-diversity to protect sensitive personal data. Furthermore, the developed strategy ensures that the shared e-health data does not reveal or isolate the original data before moving to the Hadoop distributed file system (HDFS) [5] .

In a study, two solutions are offered that protect user privacy in parking recommender systems while analyzing the past parking history utilizing k-anonymity (anonymization) and differential privacy (perturbation) techniques [6] . The k-anonymity mechanism specifically creates an anonymized database from an original parking database that contains users’ parking information. The users are indistinguishable in both methods due to differential privacy, which also perturbs the Laplace mechanism’s query response. It is now possible for customers to receive parking place recommendations while maintaining their privacy due to experimental findings on a data set built from actual parking measurements [6] .

In order to use Fully Homomorphic Encryption (FHE) schemes, a presented proxy re-ciphering as a service employing traditional methods such as threshold secret sharing, distributed semi-trusted proxy servers, and chameleon hash function [7] . The effectiveness of the developed strategy is analyzed using real-world data. Furthermore, the strategy’s security characteristics are also analyzed over the general cyber threats, ensuring that the developed method is a sensible, scalable, and easy-to-use strategy for the long-term prevention of sensible data [7] .

A method to develop a productive anonymous algorithm to maintain the privacy of private data by the data owners [8] , privacy frequently occupies a chief role in data mining strategies, and a number of anonymous algorithms introduce privacy in digging data. However, there are limits to privacy protection, and hence a novel strategy of Efficient Anonymous Algorithm (NEAA) is introduced. At first, the entity process’s raw data and the data are preserved in the database. Then, the sensitive data is evaluated using the PCA-based Attribute selection algorithm. The hiding process introduces a novel algorithm based on Anonymous (NBA). Finally, anonymous data can be produced as a result [8] . The performance of the developed system is analyzed and found to provide enhanced performance compared to conventional systems [8] .

A presented a privacy-preserving protocol based on keyword search with security for EHR system suggested [9] . Through the use of a keyword search in the cipher text, which is then once more re-encrypted by the cloud using the re-encryption key created by the patient, this method can quickly identify the history of health records that are related. It guarantees that confidential information cannot be disclosed by unauthorized users. An entity-based access control system ensures that only the intended data requester has access to the patients’ medical records [9] .

A model to recognize how the connection of records can assist in developing the entire patient profiles and thus adjoin importance to the conventional healthcare systems was presented by [11] . The usage of data anonymity shows how privacy is fulfilled by specifying the knowledge about the background and further limiting access to factual data. A semantic strategy counting policy formalization, compliance examination, and knowledge discovery was carried out to prevent the risks related to privacy with arbitrary linkages [11] .

The tuple partitioning approach is used in an effective quasi-identifier index-based architecture for privacy preservation over incremental databases on the cloud design [12] here the merging of columns generates anonymous information for the user. In addition, a packetization strategy is applied to enhance the effectiveness of privacy-preserved information further. The outcomes of the method over a real-world database have proved that the developed model effectively preserves the incremental database of large volumes compared to conventional strategies [12] .

A blockchain structure for managing electronic health records (EHR) could provide the patient with rights and control over the EHRs. The Ancile structure also exposes a blockchain system that achieves a higher degree of decentralization while admitting some nodes as having greater power. According to the evaluation, it is very unlikely that all the data will be covered while maintaining a usable and interoperable model. Ancile, however, still offers a significant amount of data integrity and privacy protection due to the use of smart contracts to partition the data [13] .

A blockchain-based medical data preservation scheme (DPS) ensures the primitiveness and verifiability of EHRs while preserving privacy for the data owner [9] . After receiving authorization from the data owner, the data requestor uses the expected keyword from the data provider to analyze the expected EHRs through the EHR consortium blockchain and obtains the re-encryption cipher text from the cloud server. To comprehend data security, privacy preservation, and access control, the technique largely uses searchable encryption and interim proxy re-encryption. Additionally, proof of agreement is developed as a compromise method for consortium blockchain to guarantee the availability of the system [9] . In order to demonstrate the effectiveness of the proposed scheme in terms of computing efficiency, the cryptographic primitives were also emulated and the planned scheme was implemented on the Ethereum platform [9] .

They suggested the blockchain-based interoperability problem in EHR and determined the number of cross-blockchain-based EHR storage strategies [10] . Initially, an EHR privacy-preserving cross-blockchain strategy was presented based on Polka Dot chain technology for EHR circulation among several private blockchains of various hospitals. The issue of removing EHR data from each hospital’s private blockchain is solved by RaaS. Using the developed mechanism constructs it beneficial and effective for the doctors to access the data and for patients to remove the unwanted EHR data when they visit various hospitals, which acts a significant function in the EHR privacy-preserving area and in contravening the remote islands of sharing medical data [10] .

The blockchain-based model was developed for the management of EHR on a distribution network to ensure the eventual privacy of patients’ health records, providing control to the patients to monitor the access of data by others through the developed method [15] . Gathering and organization of data into Big Data provide a huge possibility of varying the healthcare viewpoint, like in personalization care of patients, the discovery of the drug, the efficiency of the treatment, enhancement in clinical results, and the safety management of the patients. Furthermore, the blockchain offers a proposal for which the EHR of patients is preserved without tempering or any attacks. Then to ensure ultimate isolation and control over access to an EHR sheet on the blockchain, a channeling strategy ensures that the patients accept the entities only within a distributed network to completely access the data [15] .

In order to protect patient privacy, the Privacy-Preserving Optimization of Clinical Pathway Query (PPO-CPQ) system uses a safe clinical pathway query on cloud servers for e-healthcare., and the sensitive data corresponding to the hospitals, like treatment, expense, and medication. Under this strategy, a secure and privacy-preserving sub-protocol is initially designed, which constitutes a comparison of privacy-preservation, selection of privacy-preserving stage, and so on, to assure the privacy of the e-Healthcare system. The query is then safely executed using the greedy algorithm, and the system’s efficiency is increased by using the Min-Heap technique [16] .

Big data privacy has been successfully preserved using an algorithm known as Grey Wolf Optimizer-Cat Swarm Optimization (GWO-CSO). The generated model, which is used to create the k-anonymization database in which k numbers of duplicate records are created within the real database [17] , is obtained by changing the update rules of GWO in the presence of the CSO algorithm. In order to provide safe data transmission over to end users, the newly created technique is used in the k-anonymized database to conceal the sensitive information relating to the data owners. The created technique ensures the k-anonymization phenomena by producing the parameters required to build the k-anonymized database ideally [17] . The privacy and utility metrics consider evaluating the fitness of the solutions obtained using the developed algorithm [17] .

Mandala and Rao presented the privacy preservation model for private healthcare data [18] . The designed model mainly concentrated on introducing an effective sanitizing model to conceal the sensitive information of users. A key was framed and selected optimally using the Adaptive Awareness Probability-based Crow search algorithm (AAP-CSA) approach to conceal the sensitive medical data. Moreover, the designed strategy was analyzed in terms of various attacks with different algorithms to show the effectiveness of the designed method [18] .

A privacy-preserving chaos-based privacy-preserving encryption model was designed by to protect patients’ privacy [19] . The developed model could protect the images of the patients from a cooperative broker. To secure the essential frames of the data gathered from the wireless capsule endoscopy strategy specifically, a quick probabilistic model was devised and a prioritization method was used. The created model produces encrypted images that are random in nature, which improves computational efficiency. The created methodology also entails processing medical data without any leaks, and protecting patient privacy by allowing only authorized users to access the data [19] .

Yang presented a practical and privacy-preserving prediction scheme to predict disease risk using e-healthcare, named EPDP [20] . In contrast to the current approaches, the created EPDP successfully completes the two steps of disease risk prediction, such as disease model training and disease prediction, while ensuring improved privacy preservation. Notably, a cryptographic approach is used with a super-increasing order to efficiently collect each disease’s symptoms throughout the disease model training phase. Results from the disease risk prediction phase were evaluated using the bloom filter approach.

Chamikara modeled a solution to maintain data privacy in significant data distribution and evaluation strategies, a challenge in smart cyber-physical models [21] . The developed algorithm called SEAL is used to preserve data privacy. The linear time complexity associated with SEAL assists in working with the continuously increasing big data and data streams effectively. The method attains increased accuracy in classification, scalability, and efficiency while maintaining enhanced privacy and higher attack resistance compared to other methods. The findings indicate that the approach is appropriate for smart cyber-physical environments, including grids, cars, healthcare systems, and homes because it can control the continuous data streams produced by sensors monitoring a single person or a group of people and processing them before sending them to cloud systems for additional analysis [21] .

The remote data integrity checking strategy used the fine-grained update for big data storage [22] . The developed strategy attains general processes of modification, insertion, and deletion at any online location level in a file with a mapping relation among the block-level update and the line-level update. The analysis of the method depicts that the developed strategy assists in privacy preservation and public verification. On the other hand, it carries out data integrity with a low reduced cost for communication and computation [22] .

To protect private information, a big data probabilistic approach based on clustering was deployed in such a way as to obtain maximum privacy and minimum perturbation. In this model, sensitive data is preserved after recognizing the confidential information from the data clusters to adjust or generalize them. The resultant database is examined to evaluate the level of accuracy of the model concerning hidden data and loose data due to reconstruction. Results demonstrate that the developed clustering-based Privacy preservation strategy in big data led to successful reconstruction [23] .

In accordance with [24] , if safe encryption is used in devices or if the internet service provider or network observers analyze the internet traffic from the smart homes connected to the IoT devices, they can gather sensitive information about the activities that take place at home. To prevent the gadgets from becoming inoperable, the user must not block outgoing traffic from their residence, an example of privacy risk outcomes associated with IoT applications in smart homes. IoT’s privacy risks and threats can be user identification, user tracking, profiling, utility controlling, and monitoring. From the privacy perspective, the threat associated with user identification is the ability of the device to distinguish or reveal the identity of the person based on acquired data like name, address, or any such personal information. Such a threat aggravates other associated threats like tracking and profiling individuals’ behavior.

The high volume of healthcare data creates a big challenge, the desire for scalable storage and support for distributed queries across multiple data sources. Specifically, the challenge is being able to locate and mine specific pieces of data in an enormous, partially structured dataset [25] . The study results showed that privacy, familiarity, and security levels affected users’ trust in using IoT devices. Due to the privacy and security concerns associated with IoT, the amount of trust in devices also had an impact on people’s perceptions of risk and their attitudes towards using it.

According to a recent study conducted IoT devices are increasingly becoming pervasive in our everyday life, so it is important to understand the underpinning privacy and security risks associated with them [26] . These risk factors result in attacks by cyber attackers faced by consumers using IoT devices.

IoT has tremendous benefits but presents a reminder that IoT has security and privacy implications. The health data is susceptible, and its granularity poses significant challenges to the anonymization of personal information and thereby exposes consumers to data security and privacy risks. Unauthorized people can intrude into IoT data and use them in authorized ways. Moreover, the increased reliance upon big data and IoT-based devices heightens the risk of security threats and a vulnerability point for intruders to access users’ personal information [27] . In addition to security threats, IoT devices are vulnerable due to many factors. Firstly, the IT manufacturers of IoT devices are inexperienced regarding the risks related to data security relative to hardware or software items. Secondly, the security measures like encryption also need to be fully considered, and thirdly, it is hard to update these devices periodically with security fixes.

To conduct a comprehensive survey on the topic of privacy preservation for multi-institutional data, a systematic search strategy was employed across a diverse set of journals and sources. The search primarily focuses on the following key databases and journals: IEEE, Wiley, Elsevier, Arxiv, Future Generation Computer Systems, International Journal of Advanced Computer Science, Springer, and Berkeley Technology Law Journal. Additionally, relevant proceedings from the 2018 workshop and publications in the International Journal and Elementary Education are included.

The analysis is conducted using parameter measurements that have been employed by numerous academics to demonstrate the model’s efficacy. A number of metrics are observed, including MAE, RMSE, information loss, time, computation cost, communication cost, search time in seconds, search time per identifier in seconds, communication overhead, information loss, classification accuracy, attack range, and cost function NPCR-number of pixel change rate, UACI-unified average changing in tensity tests, minimum std, average overhead bandwidth. Observations reveal that the metrics time, computation cost, and range of attacks as shown in Table 1 are frequently used by researchers. The measures that the reviewers used are explained in Figure 2.

Depending on the journal’s publication year, the analysis is done in this section. Table 2 presents a review of the output from 2013 to 2021. Figure 3 gives detailed information and represents the majority of literary works from the 2020s that are pertinent to the analysis of privacy preservation techniques.

This section evaluates the articles that have been published in the aforementioned publications and are relevant to privacy protection techniques. Table 3, shows analysis concerning of journals and their affiliated papers. Journals from

the IEEE, Wiley, Springer, elementary education, Elsevier, international journals, arxiv, and next generation computer system can all be used to get publications about privacy preservation approaches. The vast majority of publications from Elsevier and IEEE have extensive analyses and reviews, which are interpreted in Figure 4.

Based on the datasets used by various researchers, Table 4 offers an interpretation of the analysis. Several datasets, including Apache Spark, Real Word, Mockaroo, EHR, Eron Email, Medical Record, Standard Adult, and others, are assessed. Figure 5 provides a description of the datasets used by the reviewers.