This is a recording of Global CENTRA Webinar on 10 April 2018 (at 10-11am EDT). The webinar series is hosted by CENTRA Project (http://www.globalcentra.org, supported by US NSF ACI Award 1550126), headquartered at the **ACIS Lab, University of Florida.
Topic: SafeDB: Secure Database Processing
Speaker: Dr. Francisco Maia is a post-doc researcher at HASLab, one of the INESC TEC *Portugal) research units
00:00 Introducing Dr. Francisco Maia, by Dr. Rui Oliveira, INESC TEC
02:26 Presentation begins
03:02 Context
04:30 What is this important? Why use the Cloud?
05:57 The problem
07:42 Malicious entities
08:04 Possible solutions
10:25 Limitations
11:27 The SafeCloud approach
13:17 Quick note: SQL vs NoSQL
14:30 Database access control and protection; secure query processing
17:17 Privacy-aware database basics; privacy-assuring systems
19:48 Encryption
23:11 Secure multi-party protocols
24:29 Trusted hardware; privacy-aware databases; privacy-assuring trade-offs
28:09 SafeCloud DB: Mix and match privacy-assuring mechanism
30:42 The Plan; SafeNoSQL, SafeDB and d'Artagnan
38:35 Trusted Hardware databases and limitations of SGX
40:30 Summary of current work / Discussion
41:39 Future steps - AIoT and more
42:42 Q & A begins
56:36 Virtual group photo!
Abstract: The success and pervasiveness of data analytics and machine learning, under the concept of ‘big data’, stems from an unprecedented state of affairs. Simply put, there has never been a point in time where so much data was available to process and analyze. From judicious data analysis it is possible to accurately deliverable targeted advertising, improve the design of a product or develop winner business strategies. However, the sense of wonder and amazement, which followed (and still follows) the recent advancements in big data analytics, is now being seriously confronted with the impact that massive data collection and analysis has on data privacy. Naturally, a privacy-centric discourse goes against the commercially backed drive to continuously increase data collection and analysis. In addition, this commercial drive must be considered a constant and, pragmatically, to be permanent. This context opens a number of very interesting challenges either technological, social and political.
In this webinar we will cover the approach taken in the SafeCloud Project to build technological solutions for data management in the cloud with strong privacy requirements. In specific, we will cover the technology behind secure data processing in the cloud and cover the challenges that remain to be tackled.
**Advanced Computing and Information Systems Laboratory (ACIS Lab): https://www.acis.ufl.edu/
Follow us on Facebook: http://www.facebook.com/GlobalCENTRA and http://www.facebook.com/acis.lab
Twitter @GlobalCENTRA
