|
|
|
|
|
|
|
Abstract: An immense opportunity lies in the data collected through social media, as it potentially offers an unprecedented knowledge on human behaviors and the dynamics of collective action. Unfortunately, merely accessing most of this sensitive data is today's most vexing challenge. To make matter worse, it is notoriously difficult to extract value from it, especially when mobility and social proximity are involved. Storing and processing massive amount of proprietary data in a large computing center hence remains de facto the standard of social networking. This talk proposes a different paradigm: social networking services where multiple devices, possibly under the control of different authorities, interact through local transactions. After browsing motivations and challenges to progress towards such architecture, we will propose elements of answer to the following questions: "Can data mining be run through local transactions?" and "Can the economics of personal data be reinvented?"
|
|
|
|
|
|
|
|