Countering Terrorism On Social Media With Big Data Analysis
ABSTRACT
Terrorism and violence are used by groups and individual to disrupt the normal course of events. This research provides an empirical model approach to detecting terrorist. The modeling method is an empirical formula for measuring individual actor’s level of involvement in terrorism crime. Terrorists are gaining ground virtually in the most nation of the world. This has become a global challenge as the attacks have recorded millions of deaths of innocent people, incapacitated people, and destroyed properties. In addition, the attacks rendered most communities and countries economically dormant and create a complex crisis among the people. In the research approach, the mathematical model developed is used to detect person’s involvement in terrorism. Furthermore, the mathematical model result can be used to predict or calculate the outcome of the dependent parameter from the varying independents. The overall test which determined the levels of optimization and improvement. Thus, I recommend that people’s profiling analysis has a significant contribution to terrorist detecting if integrated into the system as a solution to terrorist attacks.
KEYWORDS
Big Data, Data Mining, Social Network Analysis, Natural Language Processing, Big Data Integration, Conclusion, References.
1. INTRODUCTION
The government of the United Kingdom, for example, has a very explicit definition of terrorism which has been set out in its Terrorism Act 2000.
Terrorism is defined there as: ‘The use of serious violence against persons or property, or the threat to use such violence, to intimidate a government, public, or any section of the public for political, religious or ideological ends. According to the US’s Federal Bureau of Investigation (FBI) which defines terrorism as: ‘The unlawful use of force or violence against persons or property to intimidate a government or civilian population’. The dictionary definition for terrorism is “the unlawful use of violence and intimidation, especially against civilians, in the pursuit of political aims,” which describes it as illegal and, hence, wrong by the prevailing norms. . Since the commercial availability of the Internet in the late 1980s , users of computing devices are now connected affecting many aspects of society. Although the Internet add its own significant value, but the Internet still has opened loopholes and new forms of access for miscreants to exploit or to perform their illegal efforts with greater ease. The Internet also gave rise to a new definition of social interaction through the extended use of social media. As social networks are less than two decades old, all earlier mediums of socialization combined did not create such a wide audience with personalized experiences, largely due to Artificial Intelligence and Machine Learning methods.
1.1 Causes Of Terrorism
The causes of terrorism may either be socio economic, political and religion.

1.2 Types Of Terrorism
According to google, we have five types of terrorism which varies from definition to definition viz:
- State-Sponsored terrorism, which consists of terrorist acts on a state or government by a state or government.
- Dissent terrorism, which are terrorist groups which have rebelled against their government.
- Terrorists and the Left and Right, which are groups rooted in political ideology.
- Religious terrorism, which are terrorist groups which are extremely religiously motivated and
- Criminal Terrorism, which are terrorists acts used to aid in crime and criminal profit.
1.3 Examples Of Terrorism
We have a whole lot of terrorist which varies from country to country with different aim’s and ambition. I will only mention few out of all, which are Al-Qeeda, Boko Haram and so on…
- Al-Qaeda has mounted attacks on non-military and military targets in various countries, including the 1998 United States embassy bombings, the September 11 attacks, and the 2002 Bali bombings.

2. Boko Haram is a jihadist terrorist organization based in northeastern Nigeria, also active in Chad, Niger and northern Cameroon.

2. DISCUSSION OF FINDINGS
In the present information age, enormous amounts of data are created with exponential increases in recent years. With social media, huge volumes of data of different varieties are generated rapidly. Dealing with all this data is challenging as traditional methods of analysis such as guessing, constructing hypothesis and testing with data based experiments do not perform well because of the sheer volume and variety of data, hence new methods uncovering the insights in data must be devised. Research in Big Data is considering new techniques with the core challenge of how to process data to extract useful information.Computational resources seem to be second runners-up in this race with Big Data leading the way. Classic learning and intelligence methods also fail to perform well on large data set, and new “Deep Learning” techniques are being devised to take advantage of the available information. Interpreting natural language comes easily to humans. This poses another challenge for the processing of Big Data produced by social media sources. Complex graphs must be evaluated to extract useful information. Uncovering these relationships and making sense of them is critical to identifying terrorist groups, their recruitment strategies, and their information dissemination methods. Complex graph analysis techniques like spectral clustering, information maps should be employed for this.
Extracting meaningful or useful information and curbing terrorist-related activities require dealing with Big Data Analysis, natural language processing and network analysis. In the next section, I propose strategies for dealing with these problems.
3. METHODOLOGY
There are many strategies to deal with this problems but i propose the following strategies to address the three problem domains identified with the aim of combating terrorism using the latest technology.
3.1. Natural Language Processing
Making sense of natural language is fundamental in identifying malicious information. Natural language processing is the automatic analysis and representation of human language, which enables computers to perform a wide range of natural language-related tasks. [1] With the availability of large amounts of data, deep learning methods can be used to employ multiple processing layers to learn hierarchical representations of data. [2] A variety of model designs and methods have blossomed in the context of natural language processing.[3] Making use of the available language data, we can employ deep methods to sieve out terrorism-related information. This approach involves making semantic sense of the text and identifying underlying patterns to characterize terrorist-related information. Crawling social media sites and using natural language processing methods to analyze the content against keywords using state-of-threat matching methods, such as distance-based matching.
Natural language processing methods based on keywords and known entities are also useful for preparing and refining a continuously evolving dataset of terrorist groups, organizations, and linked people. This dataset is a prime resource for further learning and processing using Big Data analysis methods reviewed above as well social network analysis methods already discussed.

3.2. Data Mining on Reduced Data
Extracting information on terrorism and its allies from Big Data requires processing large amounts of data, which is sometimes not feasible due to limitations on computational resources and time constraints. Innovative methods are required to reduce the data while preserving information content, such as the following data reduction methods.
- Reduce the velocity of data streams before entering into storage (pre-processing). In a specific use case from the proposed algorithms show that data reduction performs effectively, and the memory requirement is reduced from 3 TB to 300 GB of RAM.
- Data sampling is useful when data sizes become too large to practically deal with the entire dataset simultaneously and has been used extensively in data mining applications. Sampling techniques include simple random sampling, stratified sampling, systematic random sampling, and cluster sampling.
- Dimension reduction techniques based on clustering, map-reduce implementations of existing dimension reduction methods, feature selection techniques, and fuzzy logic implementations. PCA, SVD, eigenvalue/eigenvector decompositions. Data mining on the reduced set with clustering to gather specific data into groups can be helpful in segregating terrorist groups. Grouping terrorist data into homogeneous classes or clusters can provide a comprehensive understanding of terrorist behaviors, while predicting the likelihood of terrorist activities in a reasonable amount of time. The process of using data mining involves the following steps:
- Identify useful features in the data.
- Appropriately apply the data mining strategies of classification, clustering, association, sequential pattern discovery, and regression to understand and predict terrorist activities.
- Use data mining algorithms to search for patterns in the dataset to identify additional malicious activities.
- Establish domain (terrorism and allied activities) understanding with relevant prior knowledge and identification of end-user goals.
- Build a dataset using natural language methods, as described in the next section, by using known keywords and entities related to terrorism and terrorists.
- Pre-process data for reduction and cleanliness.
3.3. Social Network Analysis
With a dataset in place and a continuous stream of available data from social media, data mining methods can extract useful information. However, this processing can be further improved using social network analysis techniques to unravel any underlying structures, relationships, and associations. Social network analysis is a collection of techniques that support statistical investigations on the patterns of communication between groups. Social scientists use these to analyze connected groups, and they form a basis of techniques for situational awareness and decision making in law enforcement applications. A linkage map of terrorist organizations can be created using social network analysis from which a frequency of co-occurrence of names of organizations can be used as a basis for inferring the intensity of the links. Concepts from graph theory play a pivotal role in network analysis, such as how an adjacency matrix will reflect the closeness of organizations. In addition, graph theoretic concepts of centrality and between, provide further insight into the operation and structure of terrorist organizations.
The methodology for studying violent groups is broken down in the following, as suggested:
• Mapping the group with characteristics of parallel ties, symmetric/asymmetric, negative/positive, and the quality of ties, including measurement of time spent together, recurrence of communication, and size of associations.
• Division of power within a group, including progressive gatherings, actors having greater number of ties, thick structures or a heap of associated subgroups situated in vital areas. Proportion of status or centrality measures, idea of impact as an element of actor’s significance.
• Structure and subgroups, including levels of cohesion and degree hierarchies, balance between efficiency with either a low number of repetitive connections or a high level of group centrality.
• Robustness and survivability with high density and a large number of redundant ties.
This framework leads to a deep analysis of terrorist groups and their modus operandi. Implementation of this is made possible using tools such as NetworkX and SNAP.
3.4 Physics Research Approach
Linear regression analysis (least squares) is used as data fitting techniques. Application is made to two experiments: (a) Fletcher’s trolley and (b) Hooke’s law.(a) Using Fletcher’s trolley: Newton’s second law which states that “the rate of change of momentum of a body is proportional to the force causing it and takes place in the direction of that force”.
i.e. (mv — mu) / t ∞ F or m(v — u) / t ∞ F ……..equation (2)
which gives ma ∞ F. i.e to show that the acceleration of a body is proportional to the applied force and inversely proportional to the mass of the body.
(b) Hooke’s law: in simple terms says that strain is directly proportional to stress. Mathematically, Hooke’s law states that; F = -KX …………….……………………equation (3)
where, x is the displacement of the spring’s end from its equilibrium position (a distance, in SI units: meters); F is the restoring force exerted by the spring on that end (in SI units: N or kg·m/s2); and, k is a constant called the rate or spring constant (in SI units: N/m or kg/s2). Least squares will extract information from raw data in a very precise way. Let consider a simple example drawn from physics, a spring should obey Hooke’s law which states that the extension of a spring is proportional to the force, F, applied to it.
3.5 Research Design
In the conceptual framework approach to this research, values of numbers were assigned to each variable to form a dataset. According to Sarah Bouslaugh and Paul Andrew Watters, “Before someone can use statistics to analyze a problem, one must convert information about the problem into data. That is, one must establish or adopt a system of assigning values, most often numbers, to the objects or concepts that are central to the problem in question. This is not an esoteric process but something people do every day”.
3.6 Data Collection
To obtain a dataset from each variable of people’s profiling such as name, passport photograph, biometric, bank details, residential address, nationality, work address, past education, religion, marital status, telephone number, etc.
3.7 Data Analysis Techniques
In each of the different design model of people’s profiling, a regression analysis technique was applied using statistical software.
TERRORISM RESEARCH CLASSIFICATION
The current terrorism research classification can be the view from information and communication technology-based solution approach such as; internet, software, computer network, surveillance camera, social network analysis and satellite imagery. Others include a database, cloud computing, data mining, robot, CCTV, expert system, mobile phone, cryptography, digital-forensic and biometric.

Big Data Integration
The term big data refer to the technology which includes the tools and processes, that an organization requires to handle the large amounts of data and storage facilities. The term is believed to have originated with web search companies who needed to query very large distributed aggregations of loosely structured data. Distribution of data and workload over multiple servers was the key to handle massive data volume efficiently. An example of big data might be petabytes (1,024 terabytes) or Exabyte (1,024 petabytes) of data consisting of billions to trillions of records of millions of people from different sources (for example contact center, web, sales, mobile data, customer social media, and so on). The data is typically loosely structured data that is often incomplete and inaccessible.
Data is distinct pieces of information, usually formatted in a special way. Data can exist in a variety of forms as numbers or text on pieces of paper, as bits and bytes stored in electronic memory, or as facts stored in a person’s mind. In many occasion, people have used the word data to mean computer information that is transmitted or stored. But, data is the plural of datum, a single piece of information. In practice, however, people use data as both the singular and plural form of the word, and as a mass noun (like “sand”). The integration of this huge data sets is quite complex. There are several challenges one can face during this integration such as analysis, data curation, capture, sharing, search, visualization, information privacy, and storage. There are several challenges one can face during this integration such as analysis, data curation, capture, sharing, search, visualization, information privacy, and storage. The core elements of the big data platform are to handle the data in new ways as compared to the traditional relational database. Accuracy in managing big data will lead to more confident decision making. In today’s terrorist challenge, there is a need for modern big data integration platform, from the local level to state level and up to national level. Thus, Integration is the inclusion or incorporation or use of supportive material or tools that could enhance performance as applied in the context.
4. CONCLUSION
In this paper, I presented the methods used by terrorists to spread their messages using social media. It is understood that containing terrorism-related material on social media is critical. The proposed strategy includes the use of natural language processing to build and expand a dataset by looking for terrorist related data on social networks, the reduction of the data using data sampling techniques, the use of data mining methods on reduced data to identify the patterns and extract useful information and the use of social network analysis to uncover the associations and relationship between individuals and terrorist groups, their structure and their modes of operations. Thus, I recommend that people’s profiling analysis has a significant contribution to terrorist detecting if integrated into the system as a solution to terrorist attacks.
4.1 ACKNOWLEDGEMENT
I would like to express profound gratitude to my friends in the Lagos State Polytechnic, Yaba College Of Technology, University Of Lagos in the department of Computer Sciences, on this study for their valuable comments and suggestions that improve the presentation of this research.
5. REFERENCES
[1] Terrorism definition, Available at https://en.oxforddictionaries.com/definition/terrorism
[2] History of Internet, Available at https://en.wikipedia.org/wiki/History_of_the_Internet
[3] P. Arie and P. Ami, “Social Network Analysis in the Study of Terrorism and Political Violence” Southern Illinois University Carbondale, OpenSIUC Working Papers 2010
[4] Md. S. Hossain, “Social Media and Terrorism: Threats and Challenges to the Modern Era” South Asian Survey, Vol 22 (2), pp. 136–155, 2018
[5] B. Bender, “SOCIAL MEDIA AND POLITICS: THE 2016 US PRESIDENTIAL ELECTION”, BrandBa, 2017
[6] Charlie Winter, “Fishing and ultraviolence: So-called Islamic State is known for its brutality. But it’s also hooking people in far subtler ways,” BBC, August 1, 2015. Available at: https://www.bbc.co.uk/news/resources/idt-88492697-b674-4c69-8426-3edd17b7daed
[7] B. Aparna. “Social network analysis of terrorist organizations in India”, Institute for Defence Studies and Analysis 2005.
[8] H. Bronnimann, B. Chen, M. Dash, P.J. Haas, and P. Scheuermann, “Efficient Data Reduction with EASE,” Proc. ACM SIGKDD, 2003. 42 Computer Science & Information Technology (CS & IT)
[9] Weinstein M et al (2013) Analyzing big data with dynamic quantum clustering. arXiv preprint arXiv:1310.2700.
[10] M. Trovati “Reduced topologically real-world networks: a big-data approach” Int J Distrib Syst Technol (IJDST) 6(2):13–27
[11] E. Cambria and B. White, “Jumping NLP curves: A review of natural language processing research,” IEEE Computational Intelligence Magazine, vol. 9, no. 2, pp. 48–57, 2014.
[12] T. Young, D. Hazarika, S. Poria,& E. Cambria, “Recent Trends in Deep Learning Based Natural Language Processing [Review Article]”. IEEE Computational Intelligence Magazine, 13, 55–75, 2018
[13] P. Svenson, P. Svensson, H. Tullberg, F. Ledningssystem, F. Ledningssystem, F. Ledningssystem, “Social Network Analysis And Information Fusion For AntiTerrorism”, In Proc. CIMI, 2006
[14] H. Aric, S. Pieter, & S Chult, Daniel. “Exploring network structure, dynamics, and function using networkx”. United States, 2008.
[15] J. Leskovec and R. Sosic, “SNAP: A General Purpose Network Analysis and Graph Mining Library”, CoRR, 2016
[16] Q. D. Truong, T. Dkaki, Q. B. Truong, “Graph Methods for Social Network Analysis”, 2nd EAI International Conference on Nature of Computation and Communication — ICTCC, pp-276–286, 2016
[17] R. Kitchin, “Big Data, new epistemologies and paradigm shifts”, Big Data & Society, Sage Publications, April, 2014.
[18] J. Cała, P. Missier, “Selective and Recurring Re-computation of Big Data Analytics Tasks: Insights from a Genomics Case Study”, Big Data Research, Vol 13, pp 76–94, 2018
[19] H. Estiri, B. A. Omran, S. N. Murphy, “kluster: An Efficient Scalable Procedure for Approximating the Number of Clusters in Unsupervised Learning”, Big Data Research, Vol 13, pp 38–51, 2018
[20] D. Lakshmipadmaja, B. Vishnuvardhan, “Classification Performance Improvement Using Random Subset Feature Selection Algorithm for Data Mining”, Big Data Research, Vol 12, pp 1–12, 2018
[21] K., Savita. “Impact of big data and social media on society”. Global Journal for reseach Analysis. Vol 5, pp 437–438, 2016
[22] C. Debas, “Big data analytics for exploratory social network analysis”, International Journal of Information Technology and Management, Vol 16(4), 2017
[23] M. William, C. Campbell, K. Dagli, and C. J. Weinstein, “Social Network Analysis with Content and Graphs”, Lincoln Laboratory Journal, Vol 20(1), 2013.
