By Jose Luis Agúndez (@CiberJos), Big Data Innovation Program
20 March 2014: In the current context of economic downturn, the promise of Big Data to fuel growth by creating new value out of existing data is still to be realized. This is despite efforts to foster it with new regulatory frameworks by policy makers, such as EU Commissioner Neelie Kroes, as described in her talk around “The Economic and social benefits of big data”.
One course of action for Big Data to gain traction is proposed by Alex “Sandy” Pentland, MIT Human Dynamics Lab director, and world-class thought leader in Big Data and Personal Data (watch him in a video at the end of this piece and here). He defends the need to build trust and transparency mechanisms around personal data, which guarantee that people’s privacy preferences are respected. In a context of higher trust, companies would be allowed to find insights in the data, creating value propositions that have personal, “social good” and business applications, while being transparent on the use of data.
Quoting Mr. Pentland during his talk on Digital Confidence at Telefonica Digital (found here) he describes it as follows: “We need a framework where people know what they are getting into, can control it, and you can audit it to know that it is safe and the right thing has been done”. He then explains that there will always be risks as well as rewards, and it is a personal choice to find the right balance between opt-in and anonymity. In order to explore openly this space, Telefonica Digital organized last year a Digital Confidence discussion panel with top names such as Ronan Dunne, CEO at Telefonica UK, Christopher Graham, UK Information Commissioner, together with Sandy Pentland and several panelists.
Regarding value creation vs privacy, Sandy states: “Privacy is the wrong debate. Privacy is one type of service that you could want, but what you really want is value for your data, and your want it to be safe, secure, and under your control”. Telefonica Digital has explored the value creation angle with products such as SmartSteps for Retail and Transport which was launched in the UK last year, but also with the “Datathon for Social good”, where insights extracted from mobile location data and London’s “open data” (i.e.: road incidents) provided by the UK Open Data Institute, were combined on new value propositions, as a living experiment towards a “Data Sharing Economy”.
Recent industry analysis might help understand additional causes for the cautious pace at which data is shared by companies. It turns out that 44% of companies don’t have formal Data Governance policies. This is one of the key findings of the newly released “2013 Data Governance Survey” conducted by Rand Secure. The report concludes with some recommendations very relevant to Big Data, to set companies in the right path to allow a better management and e-discovery of their data, to enable extracting new value out of it:
- Organizations should develop a formal data governance policy: “This survey shows that continually working to improve your policy can only increase the value your organization derives from its data.”
- Consider new technologies to help reach your Data management goals: With the growth of big data, its governance should be reliable, scalable, and efficient.
At this point, the question is how to articulate solutions to the problems identified: on the one hand, how to create a framework of transparency and trust around Big Data and Personal Data, and on the other hand, how to help companies raise awareness on management and e-discovery of their data assets so that brand new value propositions can be built.
As of today, there is no full solution to the trust problem, although some efforts seem to be pointing in the right direction. Initiatives like O2 UK Digital Confidence have led Telefonica innovation teams to explore a “Data Locker” concept in which a personal data dashboard allows users to visualize its own data (eg. the real social network of those you talk and text with), but also have opt-in choices to share bits in exchange for new views of aggregated data from all opted-in users, or access new services and rebates exclusive for top data-sharers.
A team of researchers at Barcelona UPC University are exploring with Telefonica researchers an interesting concept called “Objects as a Service”, which would encapsulate data, for instance Personal Data, in a form of “digital enclosure” comprising the data locked together with its access control rules and ways to handle it, or functions, in a way that it cannot be read or modified if the reader does not have the right permissions. It is still early research now, but it might help solve part of the problem. Other technology that focuses on preserving data, not as research but commercial grade, is Guardtime’s KSI which creates a “cryptographic watermark” to verify the integrity, time and origin of the data. With this, any attempt at modifying data would be detected in real-time.
Regarding governance of data, security is a key topic, especially in light of recent security scandals around theft of credit card data (e.g. Target). This is so severe that security experts point out that companies should “assume that cyber criminals are already inside the network and should be stopped from accessing the most important parts – for example, customer data and intellectual property”. To that end, Juniper Networks now offer an “intrusion deception” technology to trick attackers into fake data, also known as a honey-pot.
Another key aspect of Data governance is to enable a “Data Sharing Economy”, so that once a trust and transparency framework is in place, the data assets of a company can be shared in the data marketplace to finally monetize them. This is best exemplified by the growth of Millennial Media Exchange and similar marketplaces where offer meets demand, in a similar way to how the financial stock markets work.
So future Big Data Governance solutions would have to string it all together: a powerful dashboard for personal data allowing transparent opt-in choices (like Telefonica’s concepts on a “Data Locker”), encapsulation of data for additional security (like “Objects as a Service” would do if it were a commercial product), honey-pots for cyber criminals (like Juniper’s) and if thieves get past those, means to detect and trace any alteration of the data (like KSI would do), and finally a Data Marketplace to foster trading of data and insights (like MMX).
It might take some time and a few success stories, but if a trust environment is offered to people providing full transparency and control of the way their personal data is used, and companies mature their Data Governance policies, the “Data Sharing Economy“ will grow beyond the existing Open Data boundaries and realize an Economic growth that could be similar to what the Industrial Revolution meant in centuries past.