Big Data Applications - Big Data: Related Technologies, Challenges and Future Prospects - page 71

structure. Such structure with close relations among internal individuals but loose
externally relations is also called a community. The community-based analysis is
of vital importance to improve information propagation and for the research on
interpersonal relation analysis.
The U.S. Santa Cruz Police Department experimented by applying data to
conducting predictive analysis. By analyzing SNS, the police department can
discover crime trends and crime modes, and even predict the crime rates in major
regions [ 57 ].
In April 2013, Wolfram Alpha, a computing and search engine of the U.S.,
studied the law of social behaviors of users by analyzing social data of more than
one million American users of Facebook. According to the analysis, it was found
that most users of Facebook fall in love in their early 20s, get engaged when they
are about 27 years old, get married when they are about 30 years old, and have
slow changes in their marriage relationship between 30 and 60 years old. Such
research results are highly consistent with the demographic census data of the
U.S.
Global Pulse conducted a research that revealed some laws in social and
economic activities using SNS data. This project utilized publicly available
Twitter messages in English, Japanese, and Indonesian from July 2010 to October
2011, to analyze topics related to food, fuel, housing, and loan. The goal is
to better understand public behavior and concerns. This project analyzed SNS
big data from several aspects: predicting the occurrence of abnormal events
by detecting the sharp growth or drop of the amount of topics; observing the
weekly and monthly trends of dialogs on Twitter; developing models for the
variation in the level of attention on specific topics over time; understanding
the transformation trends of user behavior or interest by comparing ratios of
different sub-topics; and predicting trends with external indicators involved in
Twitter dialogues. As a classic example, the project discovered that the rice price
follows the change of food price inflation from the official statistics of Indonesia,
by analyzing topics related to rice price on Twitter (Fig. 6.2 ).