Privacy-Preserving Data-as-a-Service Mashups - Essay Example

Thesis Summary Task one Privacy-preserving data-as-a-service Mashups by: Mahtab Arafat Introduction Data-as-a-Service (DaaS) is a model that gives privilege customer to gain data on demand across diverse cloud platform over the internet. However, a single DaaS provider may not be able to meet the data request demanded by the consumer due to presence of efficient data from their own servers. Thus, this brings us to the concept ‘Mashup’. Mashup is a technology of the web with the capability of integrating different web application into one. An example is ‘Trendsmap.com’, which functions by displaying areas on maps of cities where many tweets of certain types of tweets are coming from on a global platform. This is possible since it obtains data from Google maps and tweeter and thus utilizes them as one. This paper is a summary of a thesis by Mahtab Arafat who proposes the development of a DaaSMashup Model, which will work with multiple DaaS providers to improve efficiency of data extracted from single DaaS Provider while maintaining high standards of privacy. Privacy-Preserving Daas Mashup Privacy-Preserving DaaSMashup framework should ensure that different DaaS provider have a platform where they can share data from their database with other DaaS on a secure platform. DaaSMachup systems will be very effective where DaaS provider cannot solely provide consumer’s data requirement. The result of integration of different DaaS providers can result to provision of clients with data high levels of accuracy, efficiency, and reliability. Some of the reasons that cause reluctance to the willingness of DaaS providers to share their data include data quality, privacy concerns, and the ability to match a data request. Cloud based framework for the preservation of the DaaS mashup is critical for enhancing the security of the DaaS providers. This paper is critical in providing information on greedy algorithm to the providers’ of DaaS mashup. In essence, the paper explores the measures of securely integrating the data from many DaaS providers. These providers are critical in preserving the security and privacy of the resultant mashup data. a) Matching of data request and quality A client may be willing to pay a price for a specific nature of data however the DaaS provider may lack enough data on their database to present to the client. Even with privacy issues put to concerned care should be taken to ensure that the data quality obtained after the integration of the systems is not compromised. b) Privacy The privacy concerns about data involves the fact that private data on respondents being licked. Most organizations are against private data of their getting on to the wrong hands as this poses as a security threat. This explicit data may include the respondents Name and SSN. In the proposed Privacy-Preserving DaaSMashup framework the concepts of anonymization, adhering to privacy models and Privacy-Preserving High-Dimensional Data Mashup (PHDMashup) are utilized to ensure high privacy standards are maintained. Anonymization techniques involve presenting the data in a way that compromises personification. This will not make it possible to know which respondent provided the data that is presented. There are several ways of anonymization; these include suppression, generalization, and buketization of data. These methods involve omission of details or general encryption of the data. Privacy models: These models ensure that data is presented in way that will enable anybody to learn any extra detains about respondents than what is already displayed in the database. Through these models privacy is maintained through utilization of the notions of K-Anonymity,ℓ Diversity, Confidence Bounding, LKC-Privacy and prevention of Attribute linkage attack and Record linkage attack some level of data privacy will be ensured in the proposed Privacy-Preserving DaaSMashup framework. Privacy-Preserving High-Dimensional Data Mashup involves non-disclosure of any extra data between the DaaS providers other that what is presented in the Mashup table. Contribution of the thesis The thesis is the first of its kind that proposes for a cloud-based platform to integrate private DaaS provider’s data and use it for providing comprehensive data to consumers while maintaining high levels of quality, accuracy, and privacy. It also provided a platform to meet the data request of clients through multiple channels to where it gives a close alternative to users where the data requirements are unmet. The thesis also provides a real life test on data to prove its functionality, which is proven possible. Participants of Privacy-Preserving DaaSMashup a) DaaS providers should offer their attribute tables for system generation. Data from different DaaS providers is the main attribute for the success of this mashup. However, the publishing of the data exclusive data should be removed for privacy reasons. b) Data consumers are the participants who would like specific data analysis to be provided to them. Thus, their request for data is required. c) Mashup coordinator is the mediator between the clients / consumers and the DaaS providers. Once a consumer places a request it is the responsibility of the Mashupcoordinator to select the DaaS providers. They ensure that the provider is relevant and then use their data. In case the mashup is unable to fulfill the client’s needs then the coordinator should provide a close alternative to the consumer. Many experiments conducted on the efficiency of our algorithm provide assurance that our model can help in satisfying data privacy. Moreover, these algorithms provide security for data mining requirements as specified by the providers of DaaS, as well as, data consumers. Implementation For implementation the Privacy-Preserving DaaSMashup, there was identification of Daas providers. Computation of the total price of enforcing the mashup and the effect is has on profits of individual Daas. Finally, they construct a mashup table through PHDMashup algorithm (FTH+12). This ensures security and anonymity of the attributes table. The implementation of this project was done in Microsoft Windows Azure where the web services are deployed and Microsoft SQL as the storage of the DaaS providers. Conclusion The successful implementation of this project proposal is an efficient addition to the provision of comprehensive and quality data to consumers. This has seen the contribution that has seen the possibility of integrating different DaaS for the purpose of data generation and provision of reliable data without the necessity of compromising the issue of privacy. Notably, mashup is a useful innovation in the web that enabled integration of information from many applications into a new web application. For instance, twitter and Google Maps use Trendsmap.com website to integrate their data. This application helps Google map to gather data on all the cities with the most users from twitter. In essence, this application helps the users to gather information from multiple providers depending on the request of consumers. This emerging cloud computing service is essential in gathering information using different network protocols in the internet. The application will help in integrating new technology in serving the needs of consumers in data integration. In sum, this thesis provides a framework for consumers with a secure integrated data survey. DaaS providers can advertise after registering their data. On the other hand, the consumers can submit data mining request without the fear of data security. The role of mashup will be to determine the DaaS provider with the ability to complete data mining request. They will consider the bid price, data availability, and data quality. Notably, data quality refers to classification accuracy of the data provided. Task two Social network-based framework for users and web Services discovery by Hiba H. Fallatah Introduction A web services are running software is that give a platform for the visual interaction between machines across a network. With the development of web 2.0 and its application social networks have provided a stepping stone for the positive development in service computing. This paper provides a comprehensive summary of a thesis by Hiba H. Fallatah who is advocating for discovery in social network-base framework for both users and web services. This proposed framework would allow users and web services to discover new peers and ask them to their networks. Thus, this service will have a great impact in the contribution in the analysis of the behaviors of users and web services on the vicinity of their social networks. The positive impact on this is that it will enhance advertising since it will give an insight of how the growth of social networks could be affected. Notably, web services are software programs that are active and are designed to provide support to machine interaction using the web. It has to be restated that web technology is a critical component in the implementation of SOA. This web technology has helped in changing the design and development of software in the current dynamic and enterprise oriented applications. Advertising helps the providers in publishing web services in the discovery models. This is essential in facilitating the process for the user to find the necessary services. However, the registry model has many challenges owing to its centralized structure and use of static description of the services. This makes it difficult for the services to provide localization, detecting and searching. Motivation and the intended contribution of the thesis The main intentions that motivate the thesis to be channeled towards this direction is the fact that in the web services discovery, service delivery need to be more user oriented and there is great need for the expansion of the exposure of web services to users and their pears. With the increase in number of web services, in the current system web services are found by screening the discovery directory, that is the UDDI. The UDDI however, is not very efficient as it does not allow us to take into account the web services interactions and experiences, this is proposed to be accommodated in the proposed framework. The thesis contributes by proposing the development of a framework that will enable web services to add new peers to its network with the help of a user’s social network. This will impact in greater personalization of services as the users are more integrated in the service development. The proposed system will contribute in analyzing the behaviors of users and web services. The system will contribute in enhancing advertisement since it will have the capability of introducing itself to new users to peers of current users and web services. Web service discovery Web service discovery is important because it contributes to the ease at which the requester finds the web service that best meets their needs. Here the author advocate for three main approaches that is sufficient in web services discovery. Structural discovery involves the requirements for the requester to specify structural requirement, which are then used in to determine the most similar web service that is suitable to the users. Lexical discovery involves the generation of similarity coefficient after the removal of stop words from the descriptions through the lexical algorithm. However, this procedure has a degree of ineffectiveness and is time consuming. Semantic discovery involves the usage of formal methods analyzing service capability. Thus, the requester is determined through the procedures of computer reasoning. However, this web discovery approaches deal with web services as independent entities and fails to consider any relation to the user. For successful integration of the wed services and the social aspect it is important to consider the web services and the relation to the user and the peer as well. Establishing the relationship between the requester, peer and web service This is important, as it will give us better understanding on the various interactions of the three. For this to be established three steps can be used to determine these relationships. The establishment of the relationship between the user and the web service; Thus, we are able to determine if the relationship is implicit or explicit. Identification of social network components including the nodes and edge of the user: The social network component includes the users profile which has information and links the user to the network. Identification of the users web service characteristics. This involves the determination of their popularity, activity, market share and the degree of satisfaction the user has in relation to other users. The web service and user discovery User discovery involves the requester of web services assisting other peers in web service discovery. Web service discovery involves the cases that the web service introduces itself to peers of a client. Both methods here have some degree of effectiveness in the popularization of wed services. Therefore, an effective method is necessary in enabling users to discover their needs. Advertising has become essential in the modern world hence the introduction of social service providers. These service providers are essential in eliminating the challenges the registry model encounter. Social service providers are critical in advertising industry since they help in providers exhibiting human like social behaviors. Moreover, social service elements are critical in helping them become proactive entities. They provide tips on evaluating the service that they provide hence enhancing advertising. Conclusion The social network-based framework for web discovery is a practical activity that is seen to involve social network in web discovery. We see the importance of the involvement of social network in wed discovery. Social network provides a platform where web services can be shared. This method performs better compared to outer methods that use only the web service social networks. Moreover, this application is critical in managing scientific research, as well as, addressing the needs of business enterprise. In essence, this application provides data to the consumers with data on demand across cloud computing platforms. Notably, this framework emphasize on incorporating the social users and social web services in the discovery process. This process involves merging web service and users nodes in similar international social network. The paper explored the engineering process of the social network while considering the web services characteristics and types. This is a critical component involving users of social networks in the discovery process. In addition, this will help the web service providers to remain active in introducing themselves t the users and advertise their products in the internet. This paper provides users with a simulation results that shows that this approach is critical in providing exposure of web services. This is essential in enhancing the discovery efficient and easier. Privacy preservation in high-dimensional trajectory data for passenger flow analysis By Moein Ghasemzadeh Task 3 Introduction Notably, this paper presents methods of realizing information is secure and privacy is not altered. Preservation of information is critical to protect individual rights, property, and secrets. On the other hand, this is critical in preserving business secrets thereby preventing unfair trade practices. Advancement of information technology has enabled people to use location aware gadgets. Imperatively, these devices proffer the users’ with an opportunity to do mining and analyzing human movements. Analyzing, sharing, and storing human movement provide new threats to personal secrets and privacy. The movement of people can be termed as the sequence of locations that one visits within different time span. The traditional privacy approaches results in low quality of information in the data stored. The paper proposes the methods where anonymity of human trajectory will be shielded using effective passengers flow analysis. In essence, this process entails extracting passenger flow graph from the raw trajectory. The result is then experimented using real-life and synthetic data sets to enable the framework overcome challenges of data anonymizing. These challenges include the sparseness, dimensionality, and sequentiality The issues addressed The use of contactless smartcard has increased in the recent past with many transit companies using this cards. For example, in Taiwan, there is EasyCard; Shanghai there is Public Transportation Card; and OPUS card in Montréal. Notably, this card provides many advantages to the users by seamlessly integrating cities. In addition, the use of these cards is easy, fast, and portable compared to the cumbersome traditional systems. The contactless smartcards are demagnetized, does not require patrons to slide the card in a prescribed way, they are user friendly, and easy to use. Moreover, charging of fare depends on the age of the passengers. In essence, they are required to register personal information when purchasing the card, which is captured while transiting. Security is enhanced on the stations to ensure that passengers with valid tickets are allowed to travel. This means that the passengers with leave a trace of information when they scan their cards. This data is collected inform of the location, identity of the passage, and the time of travelling. Transit Company needs to understand that this industry is dynamic and changing as new trends emerges. Essentially, the flow of passengers’ is non-static and subject to change depending developments and uncertainties. Therefore, these companies should ensure that their services change to meet the current needs of their passengers. This means that the companies need to share this information within their departments, as well as, external transit companies for a complete analysis of passenger flow. The goal of this approach is to facilitate data mining and management of traffic. However, sharing information with third parties poses a threat to the security of personal information and privacy that cannot be guaranteed. Moreover, this thesis asserts that preservation of personal information and privacy is essential to the cardholder. On the other hand, preservation of quality information is critical for the use by the data recipient. In essence, anonymous data will be used for different data mining tasks; nevertheless, the thesis emphasize on preservation of quality information. The common mechanism used in enhancing anonymity of information includes bucketization, generalization, and suppression. Contribution to study This paper presents three main contributions: firstly, the thesis aims a preserving information quality and spatial-temporal privacy of data. This is because many past studies on trajectory data anonymization have different information requirement. In essence, they did not focus on preserving the quality of information critical in producing passenger flow. Moreover, this study provides a design of a hybrid method using both local and global suppressions. This is a critical concept since it will ensure that there is data privacy, as well as, quality information. This will help in overcoming the challenges of anonymizing sparse and high dimensional trajectory data. Lastly, the paper presents an approach of measuring similarities between probabilistic flow graphs. This is critical in evaluating the differences in the quality of information after and before anonymization. Arguably, the experimental results on the synthetic and real-life trajectory data sets concur that our algorithms is efficient and effective. In essence, this will help in addressing the challenges in anonymizing trajectory data for passenger flow analysis. This paper proposes that the anonymization algorithm have three critical stages. To begin with, there is generation of probabilistic flow chat away from the trajectory database. Then, there is identification of sequences that violated trajectory database using a sequence of suppressions. The goal of this process is to minimize the effects on the flow graph structure. Lastly, removing of violating sequence is conducted to enhance the quality of information hence enabling passenger flow analysis. However, establishing the optimal solution in suppressing privacy is difficult. Consequently, the study proposes a greedy algorithm that is able to eliminate violating sequences using sub-optimal solution. Conclusion In sum, the development on information technology requires that transport companies develop measures to curb misuse of information gathered. There is a strong tendency by the transportation companies to share information. This enhances the risk concern and privacy issues, which can lead to security breaches, should be used for wrong reasons. The tradition privacy protection approaches will be unable to secure information, as well as, seal the gaps in loss of information. Notably, anonymizing of the high dimensional trajectory data is essential in examination of the flow analysis in the transport industry. This makes it beneficial and essential for the transport companies to designate the routes with the shortest travel time. In addition, this makes it easier for the transport companies to determine the availability of transport utilities such as bus station. Developing shortest routes is essential to the companies, as well as, the passengers. In essence, this helps in reducing the distance and operation cost to the companies. Read More