مقاله انگلیسی رایگان در مورد یکپارچه سازی داده ها در اکوسیستم اینترنت اشیا – الزویر 2018

The Big data revolution and the emergence of Internet of things (IoT) has led to large-scale analyses of data generated from heterogeneous devices in various scientific and governance domains. The data processing tasks include data acquisition, fusion, aggregation and integration (Ahad and Biswas, 2017). Often, in privacy-sensitive domains such as healthcare and smart cities, individual data streams that potentially lead to a privacy threat are identified and anonymized before data processing to prevent any privacy breaches. Consider the scenario where users have been introduced to many online social networks such as Twitter, Instagram, and LinkedIn. Due to diverse functionalities, different online social network platforms attract users for different purposes such as information seeking, sharing and social connection maintenance (Shu et al., 2017). Literature on social network analysis points that social network anonymization refers to the process to replace each user’s unique identifier (example, username) with a random string, but the network structure remains vulnerable for active and passive analyses (Shu et al., 2017). In addition, network structure from different platforms may be exploited and correlated for record linkage that reveals user’s identity. Therefore, often concerns of identification, location tracking and profiling are threats posed by integration of data about users of online social networks and open data released by the government. In case of healthcare, medical records of a patient are anonymized but when these records are analysed in conjunction with demographic and behavioural data they may uniquely identify the patients and their medical conditions. This is especially critical in case the patient is suffering from depression. On one hand, the information integration tasks are important to gain insights from data analyses and on the other, they may compromise personal privacy of a subject. The autonomous nature of IoT aggravates these privacy threats (Ziegeldorf et al., 2014).