Leveraging Artificial Intelligence in Counter Terror Operations: Neural Network-Based Classification Analysis for Subversion Control

 By Col Ranjan Prabhu
0
439

The Indian Security Forces have been embroiled with Pakistani state-sponsored terrorism in J&K since the time Pakistan first used mercenaries wholly backed by its Army to invade Kashmir in 1948. The stated policy of Pakistan is to build a narrative of an indigenous uprising of the local population against perceived illegal occupation by the Indian state. Towards this end, a huge number of resources have been pumped into funding and supporting terrorism in Kashmir which includes among all other things active recruitment of local youth through the lure of money, drugs, ideology, and social media celebrity status. The biggest challenge faced by the local population and the security forces in their campaign to stem terrorism is to stay ahead of the curve by constantly identifying youth who are likely to be radicalised and weaned away by Pakistan based terror networks[1]. The timely identification of such potential candidates will allow civil society and the security forces to take measures to prevent the radicalisation of such individuals. One of the ways to identify such youth is to study data available on social media handles to understand the sentiments of individuals or groups of individuals.

While the huge population numbers in Indian states pose a major challenge for the security forces to sift through sentiment data to understand and pinpoint individuals who may be exhibiting traits of radicalisation, recent developments in niche technologies like Big Data Analytics, AI and ML-powered by the exponential increase in computational speeds of computers allows for carrying out such analysis while sifting through humongous data being generated every minute by a deluge of social media handles. News and sentiment analysis of the local population is the key to understanding the security environment on the ground. Five decades ago, fully understanding this, Franklin Roosevelt in 1941 established an office called the Foreign Broadcast Information Service, or FBIS, to be run out of the CIA. The office’s mandate was simple: translate the news from around the world for U.S. policymakers to make informed decisions. Back then there was a handful of electronic and print media that needed to be scoured for news by analysts. Today the situation has changed drastically.

In today’s world, there is a multitude of print and digital news and social media platforms that produce humongous amounts of data which would require a huge number of analysts to sift through the flow of information thus making such an organisation uneconomical and unwieldy. The evolving challenge is not only to identify the news sources which can provide the required data but also to predict future outcomes based on sentiment analysis while sifting through millions of bytes of data that are generated continuously over a 24-hour cycle. In a world where every human is interconnected with the world 24×7 and is emitting data through social media platforms, browsing habits, blogs, vlogs etc the churn rate of data is so high that it would require a huge amount of processing power to sift through and analyse this data for the valuable info. Data science integrated with Artificial Intelligence (AI) has been hugely successful in sifting through this mass of multi-source, multi-language, disparate and scattered datasets while accurately predicting events that might occur based on “tones” of these emitted data on social media and websites.

The Global Database of Events, Language, and Tone (GDELT), created by Kalev Leetaru of Yahoo! and Georgetown University, along with Philip Schrodt and others, describes itself as “an initiative to construct a catalogue of human societal-scale behaviour and beliefs across all countries of the world, connecting every person, organization, location, count, theme, news source, and event across the planet into a single massive network that captures what’s happening around the world, what its context is and who’s involved, and how the world is feeling about it, every single day[2]. This database used millions of publicly available records, robotically analyzed for tone along 1,500 dimensions, to accurately pinpoint the location of Osama bin Laden within a 200-kilometre radius of where Bin Laden was eventually found in Pakistan. The co-founder Kalev Leetaru used the datasets provided by the database to effectively predict events of a social or civil uprising such as the Arab Spring[3]. Artificial Intelligence & Machine Learning (AIML) tools are today powerful enough to capture, study and provide desired outcomes from a disparate dataset warehouse which may contain a variety of structured and unstructured data.

Within the ambit of AI, Neural Network-based classification analysis is a powerful Machine Learning (ML) based model which is being used in the civil industry like banking and finance, telecom, and insurance sectors to predict customer sentiment analysis and customer churn using parameters like age, geography, financial status, gender, use of financial products/usage of telecom services, sentiment analysis picked up from call centre calls by customers etc to predict if the customer will stay with the company or leave it [4]. This allows the sales department of the firms to strategise and timely implement measures to prevent the churn thus reducing loss of customer base. With the availability of such powerful tools in the public domain, it is pertinent that the Intelligence Community and the Security Forces invest in AIML based Data Analytics to be able to predict sentiments amongst the local population using Neural Network-based classification analysis to accurately predict the likely possibility of radicalisation and joining of militant ranks by the local youth. Such an AI-based project using Neural Networks, if developed and implemented to capture data of local population in J&K based on certain identified parameters like age, gender, economic status, place of stay, educational status, political affiliations, religious affiliations, and social media sentiments etc can carry out accurate classification of individuals to identify the probability of radicalisation or joining of militant ranks by such identified youth which can help the security forces to implement corrective measures in concert with the civil society to prevent them from becoming a victim of Pakistan based propaganda, thus saving precious human lives.

The ethical question in implementing such a project would stem from issues of privacy which would arm Human Rights groups to stall such a project due to concerns of hounding innocent local youth if prediction and classification error rates are high. This, however, can be obviated by training and testing the AI Model based on a dataset that is actuals. Application of appropriate methods to minimise errors to optimize the predictions and reduce errors while training and testing the AI Model is, therefore, the key to ensuring a high rate of success in predictions and classification. The most difficult part of building an AI project for such a task is the availability of training and testing data. However, due to five-decade-long terrorism in J&K, a huge amount of data wrt youth who joined the militant ranks due to radicalisation by Pakistan based terror group already exists. Thus, if this data is collected and collated within well thought-out parameters, it can be used with existing models available in the commercial domain of corporate industries which predict customer churn, to train, test and perfect the proposed AI Model which can then be used to predict the probability of radicalisation and joining of militant ranks by classifying local youth with minimal errors. The AI Model can then allow for on the fly sentiment analysis by sifting through known social media handles and dovetailing this analysis with the available dataset of other collected data to classify an individual to understand who may have a higher probability of radicalisation and joining of militant ranks. These predictions can also be helpful to policymakers who can utilise outputs of such a model to calibrate and re-calibrate their surrender and rehabilitation policies[5].

Decade long terrorism in J&K has scarred the population and prevented meaningful economic and social development. The abrogation of Article 370 has presented an opportunity to both the lawmakers and the civil society to infuse development in the state. The effects of hurdle free development are beginning to be seen in many areas in J&K. This has unnerved Pakistan which sees itself losing the leverage it once had over the local population. It will therefore try its level best to put in place innovative measures to recruit more and more local youth into its terror network with an aim to showcase to the global audience that the terrorism in J&K is a local uprising rather than violence perpetrated by its terror networks. Availability of a huge amount of computing power and sophisticated AI-based Data Analytics in the world today allows for the inclusion of specific AI Models in government white papers and strategy formulations[6] which can be used to give both the security agencies and the civil society an informed edge to curb the menace of terror and prevent valuable Human Resources being sacrificed at the altar of Pakistani state-sponsored terror packaged as religious fundamentalism, fanatism and indigenous uprising.

Endnotes:

[1]“Recruitment of youth into terrorism in Kashmir is a big concern for the Indian Army”, https://www.tribuneindia.com/news/j-k/recruitment-of-youth-into-terrorism-in-kashmir-is-a-big-concern-for-the-indian-army-157181, accessed on 24 October, 2021

[2]GDELT, https://en.wikipedia.org/wiki/Global_Database_of_Events,_Language,_and_Tone#cite_note-1 , accessed on March 29,2019

[3]“How the Internet Could Have Predicted the Invasion of Ukraine”, https://www.defenseone.com/technology/2014/04/how-internet-could-have-predicted-invasion-ukraine/82480/ , accessed on March 29,2019

[4]Ali Moez – “Predict Customer Churn (the right way) using PyCaret”, https://towardsdatascience.com/predict-customer-churn-the-right-way-using-pycaret-8ba6541608ac , accessed on October 26, 2021

[5]Dinaker Peri – “Rehab policy for Kashmiri youth joining terror in the works: senior Army officer”, https://www.thehindu.com/news/national/other-states/rehab-policy-for-kashmiri-youth-joining-terror-in-the-works-senior-army-officer/article32875061.ece , accessed on 26 October, 2021

[6]Brigadier Vivek Verma – “Non-Contact Warfare – An Appraisal of China’s Military Capabilities”, Pentagon Press pp 284