derives improvement suggestions, that is,
recommends words or expressions that are more
appropriate in the given situation.
5 CONCLUSIONS
Big data pundits, mainly the providers of database
technologies, allege that big data technology creates
a new generation of analytical tools that help us to
solve many problems in crime prevention, health
care, and global warming, just to name a few. There
is no doubt, big data technology is very good at
detecting correlations. However, it does not
necessarily mean that there is a plausible causation
behind every correlation detected. Even if we focus
just on correlation analysis, we have to admit that it
can get quite complex and versatile. Usually, it starts
with forming a hypothesis, that is further transformed
to an analysis model and finally to a set of instructions
for the analytic tool. The results are then checked
against the hypothesis, the model is adapted and
tested again. After a couple of iterations the causation
behind the correlation is ideally validated. We can
easily imagine that there any many and subtle
parameters to fine-tune in this analytic cycle. Hence,
there is no such thing like general purpose big data
analysis.
Big data technology can take us quite close to the
information we require. However, in many situations
we need to adapt the technology or we need additional
tools for the essential final processing step. The actual
discussion in big data bypasses this missing link in
data analysis. For a successful big data architecture,
we need a broader technology mix that reaches
beyond database, statistical, and dashboard features.
Without closing the technology gap in analysis we
cannot fully leverage big data. There is a series of
tools required that ranges from domain to domain and
from methodology to methodology. Even if we just
look at text mining, there are standard mining tools
for texts, but things like opinion mining and analysis
of customer requests or complaints, for instance, can
get very specific and reach definitely beyond mining
features. The automatic detection of positive, neutral
or negative opinions in social media about products
or service of a specific company needs to be gauged
continuously. It can be quite beneficial if the
information consumers, in this case the decision-
makers, can immediately adapt the mining features.
This article outlines the problem of the technology
gap in the context of automatic translation. With an
additional tool that enables the users to retrieve
essential expression in the right context, they can
correct diction errors that occur frequently in machine
translation. Even though machine translation does the
heavy lifting, the users need to correct its results.
Only by this final step they get the expected quality
in their translations. We consider the context-
sensitive translation memory as an example among
many, many add-on tools that will be required to
produce the insights we can expect from big data.
REFERENCES
Casacuberta, F., Civera, J., Cubel, E., Lagarda, A.L.,
Lapalme, G., Macklovitch, E., Vidal, E., 2009. Human
interaction for high-quality machine translation,
Communications of the ACM, vol. 52, no. 10, 135-138.
Feldman, R., 2013. Techniques and applications for
sentiment analysis, Communications of the ACM, vol.
56, no. 4, 82-89.
Hendershot, S., 2016. Data done right, Project
Management (PM) Network, March 1, 2016, published
on http://www.pmi.org/learning/library/data-done-
right-project-success-9989, retrieved on March 13,
2017.
ifo Institute for Economic Research, Ifo Business Climate
Index, http://www.cesifo-group.de/ifoHome/facts/
Survey-Results/Business-Climate.html, retrieved on
February 28, 2017.
Jagadish, H.V., Gehrke, J., Labrinidis, A.,
Papakonstantinou, Y., Patel, J.M., Ramakrishnan, R.,
Shahabi, C., 2014. Big data and its technical
challenges, Communications of the ACM, vol. 57, no.
7, 86-94.
Kavis, M., 2015. Forget big data - Small Data Is Driving
The Internet Of Things, Forbes, February 25, 2015,
published on https://www.forbes.com/sites/mikekavis/
2015/02/25/forget-big-data-small-data-is-driving-the-
internet-of-things/#6f1ab9cb5d7e, retrieved on March
13, 2017.
Liu, J., Wilson, Gunning, D., 2014. Workflow-based
Human-in-the-Loop Data Analytics, Proceedings of the
2014 Workshop on Human Centered big data Research,
49-52.
Marcus, G., Davis, E., 2014. Eight (No, Nine!) Problems
With big data, New York Times, April 6, 2014.
Tucker, P., 2014. The Naked Future: What Happens in A
World That Anticipates Your Every Move?, Current.
Wright, A., 2009. Our sentiments, exactly,
Communications of the ACM, vol. 52, no. 4, 14-15.
Zhang, L., Liu, B., 2011. Identifying noun product features
that imply opinions, HLT ’11, Proceedings of the 49th
Annual Meeting of the Association for Computational
Linguistics: Human Language Technologies: short
papers – vol. 2, 575-580.