The Importance of Policies in Big Data Management

Valerio Eletti :: The Importance of Policies in Big Data Management ECCS14 :: Lucca, 24 settembre 2014

COMPOLICY

ECCS, Satellite “Compolicy” Lucca :: 24 September 2014

Satellite COMPOLICY

keynote speech by VALERIO [email protected]

THE IMPORTANCE OF POLICIES IN BIG DATA MANAGEMENT


COMPOLICYAgenda

Agenda

we open the Satellite with a general survey that enables us to better contextualize the oncoming speeches of the day Complexity and Big Data

Disambiguation

Concrete examples of mega data sets

Turbulent environment

Three policy hypotheses

Temporary conclusions

The speeches of Compolicy


COMPOLICYComplexity and Big Data

What does Complexity have to do with Big Data? They have both technical and social resonances

A. Technical resonances: in Big Data, as in complex systems, we can observe:

Great abundance of elements Non-linear relationships among elements Emerging patterns Possibility of spotting clusters and hierarchies

(emergence of complexity levels) Possibility of studying the topology of the network of

relationships between data and their clusters


COMPOLICYComplexity and Big Data

What does Complexity have to do with Big Data? They have both technical and social resonances

B. Social resonances:

Big Data concern complex social relationships and complex social statuses

Mega data sets can be studied with a “memetic” approach (viral diffusion of signals)

Big Data entail the complex problem of conflicts and cooperation between public benefit and the right to privacy


COMPOLICYAgenda

agenda

Complexity and Big Data

Disambiguation







COMPOLICYDisambiguation

Let’s have a look at the terms Big Data and Privacy

A. Big Data: two different main typologies:

Mega sets of historical consistent data (that can be used for instance to train neural networks for simulations)

Composite and non-consistent mega data sets from which we can obliquely extract meaning (classic examples: forecasts based on the queries of Google on the spread of flu; previsions on stock market based on Twitter sentiment). This is the category that mainly requires clear policy guidelines in Big Data management to protect citizens privacy.


COMPOLICYDisambiguation

Let’s have a look at the terms Big Data and Privacy

B. Privacy: a notion to clarify

Past. The notion of Privacy has been changing over the last centuries: from village to city, and back to village (the McLuhan’s global one)

Present. The notion of Privacy is now in a turbulent phase of transition: for example, see how Byung-Chul Han, in his research Transparenzgesellschaft, 2012, defines our current society: of positive, of exposure, of evidence, porn, of acceleration, of intimacy, of information, of unveiling and of control, in the current post-capitalist world


COMPOLICYAgenda

agenda


Disambiguation







COMPOLICYExamples of mega data sets

Let’s see some current - extreme and paradoxical - concrete cases of mega data sets that sweep up data, and that require an accurate management of the complex problem of the conflict between public benefit and privacy:

A. Internet of things

B. Sharing our own DNA in social networks

C. Real life analytics

D. Big Data from porn



A. Internet of things Vehicle-to-Vehicle Communication Technologywill be law in the USA by 2017

http://www.safercar.gov/v2v/index.html



B. Sharing our own DNA in social networks 23andMe: the largest DNA ancestry service in the world

Https://www.23andme.com



C. Real life analytics Pathflow: “We study how customers move and behave inside a location … and we generate dashboards with objective data and analysis …”

http://www.pathflow.co



D. Big Data from porn “Pig Data” are the huge quantities of data that the Mindgeek group gathers from its Web porno sites as PornHub and YouPorn: in the year 2013 they had 14,7 billions of visitors who saw 63,2 billions of videos. Few people know that:

PornHub works on 5.000 terabyte everyday; while Facebook works on 600 terabytes…

PornHub has 1 billion visitors every month; while Tumblr has less than one fourth…

Wired Italia, September 2014


COMPOLICYAgenda

agenda


Disambiguation







COMPOLICYTurbulent environment

We must not forget that Big Data revolution is happening in a context that becomes more and more alarming as the followings items grow exponentially:

The strength of hubs that can collect digital data social networks, as Facebook or Twitter search engines, as Google cloud providers, as IBM or ChinaCache gatherers of videos/images, as You Tube or Pinterest telephone companies, as AT&T or Vodafone banks and insurance companies, as Axa or ICBC geolocation’s systems, as Geobytes or IP2Location e-commerce hubs as Alibaba, Amazon, eBay,

Taobao

…



…

The secret agreements of these scraping-data hubs with intelligence agencies

The development of semantic engines able to perceive the meaning (ref. Web 3.0, as foretold by Tim Berners-Lee)

The application of effective instruments such as soft computing and artificial neural networks

…



…

The available memory and the computing capacity, as for example:

Systems of parallel computing (artificial intelligence) as Watson IBM (that demonstrated enough intellectual abilities to win at “Jeopardize”)

Systems of quantistic computing that can decode any kind of encryption,

... and even survey systems in our subconscious, such as the one studied by CEEDS

…



…

and the remote but real possibility to reach the singularity that could mark the overtaking of the human brain by artificial intelligences


COMPOLICYAgenda

agenda


Disambiguation







COMPOLICYThree policy hypotheses

We can individuate three main approaches to Big Data management

Encrypting on-line messages

Bill of digital rights and duties

Pricing and selling our data to Web giants




oMain reference: Google policy to Chinese users

oThe policy followed by many web and mobile companies to associate the encryption to messages’ self-destruction (i.e.: Wickr, Telegram, RedPhone, TeextSecure)

oA solution admired by opinion leaders such as Snowden and Greenwald (see the speech at SXSW Festival in Austin, 2014)

oBut …




o… But there are two questions to be answered:

• Can we really trust corporations asserting that they will encrypt or destroy our messages?

• And, even if we trust them, and we subscribe to those new encrypted networks, who will protect the acquisition of our data, which is in any case done outside the Internet and the mobile networks?



Bill of digital rights and duties

oIt is a proposal referring to the origin of Western democratic tradition (see Tim Berners-Lee, The Guardian, 12 march 2014 or the video “Establish web’s principles on openness and privacy”)

o… But we have two more questions:

• Are we sure that a “Magna Charta” is sufficient to defend citizens’ right of privacy?

• And why similar initiatives such as the one proposed by Stefano Rodotà in 2007 did not lead to any result?




oThe first start-ups based on this idea are about to be realized; see for example Datacoup, that offers 8 euros per month to those who willingly allow the company to use their on-line data https://datacoup.com

oThis is a bottom-up solution (not a top-down one, as the two previous cases): a solution from the inside, characteristic of complex systems that are able to organize themselves on simple and clear laws (ref. to the Jaron Lanier, “You are not a gadget”, 2010)

oKey sentence: “The one who steals the privacy defends his privacy very well” (see “Le Scienze”, February 2014)

oBut …




o… But there is a question:

• Assuming that the bottom-up settings will work on mobile and Internet, what about the protection of data that reach big corporations and governments’ storages via banks, video surveillance, Internet of things, geolocation, health care facilities, traffic monitoring and so on?


COMPOLICYAgenda

agenda


Disambiguation







COMPOLICYTemporary conclusions

Temporary conclusions: actors

Up to now it is still impossible to foresee the solutions that will be found, dictated or accepted by single actors, or at a general rules’ level. Anyway, we can distinguish three different areas of interest:

o Area of both individual and collective data

o Area of collective anonymized data

o Area of data on actions/contacts of single citizens

…




o Area of both individual and collective data:

Public: sets of data collected by intelligence agencies of democratic or not democratic governments

Private: data collected, analyzed and utilized by private companies

…




o Area of collective anonymized data

Public: data sets (mostly open data) from which tendencies to handle traffic, epidemic, and so on are gathered

Private: data sets from which private companies can gain information for viral marketing

…




o Area of data on actions/contacts of single citizens

Public: data used with the aim to get social and political control led by intelligence agencies and governments, both democratic and totalitarian ones

Private: data used to improve a customized commercial offer by big companies managing every kind of networks



Temporary conclusions: territories

The issue of managing big data acquires different aspects of social risk depending on the intersection between

o geographical position (i.e. more or less democratic countries vs totalitarian ones)

and

o involved actors (intelligence agencies, telecom, web companies, and so on)



Temporary conclusions: a note

Commonly, in Western countries private citizens are not worried about leaving their traces in the web, in search engines, or in social networks, because of a deep-seated trust in liberal governments’ use of those traces to control and protect citizens



Temporary conclusions: policies

The solution to the problem of privacy cannot be just one, imposed from above, the same for everybody, but it has to be found by attempts, mistakes and corrections, starting from the bottom, as it always happens in biological and social complex systems


COMPOLICYAgenda

agenda


Disambiguation







COMPOLICYThe speeches of Compolicy


There are thousands of ongoing researches deepening singular aspects of Big Data managing and trying to plan a shared setting.

Among these,some of the speeches we are going to hear today are of particular importance.

So, I am now asking Francesco Niglia to give a survey of the next speeches ...




Agents reasoning under uncertainty: an extended BDI model in the JADEx framework

Forecasting transitions in systems with high dimensional stochastic complex dynamics: A Linear Stability Analysis of the Tangled Nature Model

Use of multi-level self-regulating agents

Discovering best practices in emergency response with combined physical and organizational simulation

…




…

Risk-Based Compliance and the Power Grid as Smart Territory: A Cased-Based, Complex Systems Perspective

Mutual shaping between technologies and law: e-health

Product End of Use Management Networks: ABM Approach

Spatial and Topological Diffusion of Electrical Vehicles in Urban Environments


COMPOLICY

Thank you.

You can contact me at: [email protected]

or on the site of Complexity Education Projectwww.complexityeducation.it

Thanks

The Importance of Policies in Big Data Management

Internet

terms big data

big data management

big data managementto

big datawhat

dashboardswith objective

social networks23andme

social networksc

social resonancesa