We
all know that Google and many other companies track our movements on the net to
show specific ads that they consider best fit our consumer profile. However,
most do not know how this information is used exactly. According to a recent
study conducted by researchers at Carnegie Mellon University (USA) and the International
Institute of Information Technology, part of the value judgments that arise
advertising system established by Google to many might seem undesirable.
To
conduct the research, it has created a tool called AdFisher that investigates
the segmentation of the ads displayed by Google to web pages of third parties.
The results revealed that false Internet users who Google thought were men and
seeking employment were more likely to find jobs for positions of greater responsibilities
and better paid than women.
In
addition, the study has also shown that the tool settings ads, which lets you
view and edit the interests that a company has assigned to you, does not always
provide information that could be used to target advertising.
The
reason for these patterns is still unknown, since the algorithm of Google is
still very complex. Google can use this information to target advertising to
advertisers, but these can segment your target audience based on certain
parameters and data already held. But the truth is that this discrimination is
becoming worrying, since they can also have a great influence on the decisions
of the population.
But
AdFisher is not the only tool that is designed for this type of analysis. XRay
was also launched last year to study the connections between the ads displayed
to Gmail users and keywords included in your message.
The
results of both tools cannot be seen as definitive conclusions as could be
specific exceptions. So it is necessary to design tools that can conduct
studies on a much larger scale, in order to observe how the algorithms track
users to target ads if groups of civilians and levies will be aware of the
progress they occur in the form in which this information brands try.
No comments