SOCMINT AS A PREDICTIVE ANALYSIS

TECHNIQUE
Darius COCÎRȚĂ
Academia Forțelor Terestre „Nicolae Bălcescu”, Sibiu, [email protected]
Col. Assoc. Prof. Daniel COLESNIUC, Phd

ABSTRACT
The quick development of social networks in the last 10
years has not only transformed the way society works, but
also how it can be studied. Throughout history, most of
human activity, naturally and normally, has been lost. As
people started to move bigger parts of their lives into the
online environment - where they live, what they are
studying, what political orientations they have- the normal
agglomeration of everyday life is captured in digital form.

The exponential growth of these new types of information

about what we believe, what we do and what we like, has
led to the emergence of a new type of research discipline:
social media intelligence or SOCMINT.

Keywords: SOCMINT, analysis, monitoring, data, information.

Reference elements in Social Media Intelligence

In terms of the purpose of this paper, we can define the concept of social media
analysis as the science of extracting valuable hidden information from large volumes of semi-
structured and unstructured data for the advanced knowleged of decision makers.1
In the same way, the monitoring of social media is the continuous process of tracking and
collecting information about what the public is discussing on these platforms - a keyword, a
hashtag, an event, a movement.2 Having now understood this two concepts - social media
analysis and social media monitoring - we can define a third concept, social media
intelligence, as the process that uses both social network monitoring and content analysis to
get strategic solutions.
In order to a better understaning of the role of each concept, we can look at the figure
below:

Grimmelmann, J., Virtual Borders: The Interdependence of Real and Virtual Worlds, (New York, Peer-
Reviewed Journal on the Internet), 2006, 15.
2

http://www.telegraph.co.uk/news/worldnews/11207681/How -using-social-media.html, accesed on 23.01.2019.

 Figure 1 - SOCMINT components
We have determined that the intelligence process applied in the social media is
completely dependent on the two notions explained above: monitoring and content analysis.
Using them in an objective and empirical manner, we can launch predictions on various areas
of operational and strategic interest.

The management of non-structred data

With the explosion of information resources on online platforms, organizations which

are seeking feedback in this environment are confronted with a huge amount of data. 3 We can
catalog social media data into two distinct categories: structured data and unstructured data.
Structured data refers to the type of data that is organized and displayed in a database
with a default structure like rows and columns, making it easier to work with them. Examples
include names, phone numbers, friends lists, etc.
Nonstructured data instead, are data which do not ave an structured organization. Due
to their variability and structure, unstructured data can not be handled manually. Some
examples of unstructured data are:
 Posts
 Pictures
 Comments
 Members of some communities
When monitoring social media, we see that most information can not be segmented
into fixed categories due to the complexity and variability of the content. Social media users
write about different topics in different forms, making it difficult to categorize them in a strict
manner. 4But with the growing interest for online data, new data analysis software tools have
been developed to understand such unstructured data. Such programs, with the help of
keywords and parameters, can give value and classify the data according to the areas of
interest. These programs can usually be used with license purchase, but such softwares make
analyst work much easier. Here are some of these software: Knime, Stalkscan, Zoho Social,
Hootsuite, Falcon.io etc.
3

Omand, D., Bartlett, J., Miller, C., Introducing Social Media Intelligence (SOCMINT). (Intelligence and
National Security), 2012, 35.
4

Kosinski, M., Stillwell, D. e Graepel, T., Private Traits and attributes are predictable from digital records of
human behaviour. (Proceedings of the National Academy of Sciences of the United States of America), 2016,
55.
 An important thing to be considered by organizations in different sectors that are
interested in social media resources would be investing in such software tools to extract as
easily and reliably meaning from these unstructured data for the decision staff.

Figure 2 - Social Media data analysis software

Research and knowledge based on SOCMINT

Social media research can help to understand a large number of phenomena. These
phenomena could include violence indicators, ways of radicalization, calls to aggression, etc.
In addition, this analytical approach to social platforms now offers a new law enforcement
opportunity based on the generation of operational and strategic information that could help
identify criminal activities or early warning of outbreaks by providing information about
groups and individuals.5 We will exemplify some of these practical ways that support the
predictability of SOCMINT's content monitoring and analysis:
 Chronological gathering: This is the ability to collect and group information in a way
that describes the crowning of events. Social media study has shown that while much
of social media trafficking related to an event occurred after the event, there has
always been a certain misleading content.
 Studying online communities: This would include the ability to better understand the
activities and behavior of groups. SOCMIT can track topics of interest within a target
group and how the group reacts to the external environment. This technique could
indicate the general level of anger within a suspect group, as well as their key
concerns and themes that animate group discussions.
 Individual Behavioral Research: It is applied in a legal and ethical framework by monitoring of
competent staffs on persons who have suspicious behaviors, regarding with the rules of public safety
and security. This research of the information presented in the individual's digital profile takes place
within the limits of the confidentiality measures assumed by the individuals. Studying carefully these
data is necessary because there is a certain risk that they are not true, so a contextual understanding
must be approached.6

Sumner, C., Byers, A., Boochever, R., Park, G.J., Predicting Dark Triad Personality Traits from Twitter usage
and a linguistic analysis of Tweets, (Proceedings of the IEEE 11th International Conference on Machine
Learning and Applications ICMLA), 2012, 23.
6

https://ro.ecu.edu.au/cgi/viewcontent.cgi?referer=&httpsredir=1&article=1092&context=ism, accesed on
12.01.2019.
 Such potential suggests that Social Media Intelligence plays a special role in
information structures, especially because the fact that social networks represent the interface
between the state and the citizens.

Conclusion and recommendations

The development of a social networks has a major impact on the social aspects
worlwide. We are all aware that social media platforms has had a huge impact on many
aspects of our lives, revolutionizing the way we communicate and socialize.
The opportunities that arise with the proliferation of social networking, which I have
exemplified in this paper, have led to an understanding of the need to analyze social media in
order to make informed operational and strategic decisions in an informed and timely manner.
The aspects which we analyzed in the present paper suggest that the analysis based
on social media data, gained a stand-alone position along with the other traditional ways of
collecting information and measuring the feelings and public opinions of users. Social Media
Intelligence is a process that includes identifying and collecting data and information on
social media networks, based on which the analyst can produce relevant predictive analysis.
An informational management that is fully focused on processing and studying
online resources has a predictive capacity to combat certain phenomena that can jeopardize
public safety and security.
Adopting such an information management based on social media sources, is
particularly necessary in the context of developing new capabilities that respond to new
challenges based on necessity and legality.
SOCMINT is a new way of predictive analysis which is being a real challenge to the
intelligence process because of the enormous collection of raw data, some of them false, their
dissemination and their use.

Bibliography

1. Grimmelmann, J., Virtual Borders: The Interdependence of Real and Virtual Worlds, New
York, Peer-Reviewed Journal on the Internet, 2006
2. Kosinski, M., Stillwell, D. e Graepel, T., Private Traits and attributes are predictable
from digital records of human behaviour. Proceedings of the National Academy of
Sciences of the United States of America, 2016
3. Omand, D., Bartlett, J., Miller, C., Introducing Social Media Intelligence (SOCMINT).
Intelligence and National Security, 2012
4. Sumner, C., Byers, A., Boochever, R., Park, G.J., Predicting Dark Triad Personality
Traits from Twitter usage and a linguistic analysis of Tweets, Proceedings of the IEEE
11th International Conference on Machine Learning and Applications ICMLA, 2012
5. Tomes, R., Toward a Smarter Military, Socio-Cultural Intelligence and National Security,
The US Army War College Quarterly Parameters, Contemporary Strategy and Land-
power, 2015

Web sources
1. https://ro.ecu.edu.au/cgi/viewcontent.cgi?
referer=&httpsredir=1&article=1092&context=ism
2. http://career-intelligence.com/social-media-negatively-affect-career/
3. http://arnoreuser.com/publications/
4. http://www.telegraph.co.uk/news/worldnews/11207681/How -using-social-media.html.
5. http://www.businessinsider.com/the-biggest-difference-between-google-and-facebook-
2015-4