Information Technology Reference
In-Depth Information
2.1
Predictive Analytics
Increasingly sophisticated analytical tools have spawned the field of predictive
analytics, where predictions about phenomena can be made from an analysis of data
mined from Big Data. But there are serious concerns with this predictive ability. This
has already been demonstrated in real life by the Target company case, where the
company collected data on the purchase patterns of woman customers and were able,
from an analysis of that data, to identify pregnancies and predict due dates. This led
the company in one instance sending pre-natal and post-natal material to a young
woman whose father became very upset as he had no idea the daughter was pregnant.
Similarly, MIT researchers were in 2009 able to design a program that would analyze
Facebook friendships and was thus able to predict male sexual orientation [6].
But the predictive analysis can raise issues even at the societal level. For instance,
there are concerns that it could lead to the creation of police “pre-crime” departments
that arrest individuals from decisions made on the basis of predictions of committing
a future crime. Also, predictions could lead to redlining neighborhoods in insurance
or social services coverage. As Tene and Polonetsky [4] point out:
In a big data World, what calls for scrutiny is often not the accuracy of the raw data
but rather the accuracy of the inferences drawn from the data. Inaccurate,
manipulative or discriminatory conclusions may be drawn from perfectly
innocuous, accurate data.
Concomitant with the predictive analytics problem is the automated decision-
making that data mining system often make about individuals, regarding such things
as their creditworthiness and insurance eligibility. These decisions are often made
without any human intervention and there is little opportunity for the individuals to
give their feedback or question the underlying data. Even if they wanted to give notice
to the individuals concerned before using their data, Big Data miners cannot know in
advance what they will find and thus cannot give notice of purpose for the
information and the use to which it will be put. Relying on informed consent is
likewise impossible, because data subjects cannot be able to monitor all the
correlations and new patterns that will be produced by mining their data [2].
3
Big Data and Society
Industry is not the only Big Data driver. President Obama launched a significant Big
Data research and development initiative in 2012 to bolster the study of the extraction
of information from large heterogeneous data sets. Thus, with industry, social
behavior, and government behind it, Big Data will only grow larger and the privacy
problems associated with it are going to grow not in tandem, but exponentially.
Big Data has however also exposed deep inequalities in the power balance
between the data generators and the data holders. Manovich [7] states that there are
three classes of people in the Big Data world: the ones creating data either
deliberately or by leaving digital footprints, those with the ability to collect the data,
and the few privileged ones who have data analytical expertise. He sees these groups
as forming a pyramid, with the data generators at the base, and the experts at the apex.
Search WWH ::
Custom Search