– by ORKASH Labs, Copyright: ORKASH Services Pvt Ltd

Social media has become a catalyst for civil mass movements and social unrest across the world. This includes upheaval in Tunisia and Egypt, Iranian election protests, disturbances to law and order across India in response to Delhi gang rape case in December 2012, Anna Hazare anti-corruption movement in India, the 2011 riots in London, etc. The list goes on.

This proliferation of Social media, especially through the ubiquitous mobile phone, coupled with bursting population in urban areas, poses an unprecedented challenge as well as an opportunity for Law Enforcement agencies across the world. Social media provides a powerful communication platform for organising protest and civil unrest; but on the other hand it can give government and police agencies with the means for real time intelligence, and, more importantly, the ability to intimately understand the ‘pulse and mood’ of the people; for example their reason for discontent and the underlying societal stress points of a community group. Social media also has a tremendous potential for creating accountability and governance transparency through ‘virtual’ non-intrusive partnerships between the police and the local communities.

Public Partnership for Policing – Boston Bombing

Take the case of the recent Boston Marathon terror bombing. Unlike the last time the continental United States was attacked, (11 Sep 2001, when social media sites like Facebook and Twitter were not even conceptualized), this time in Boston the social media platforms became a shared public repository for video and photos from the scene, with the people at large, as a result, becoming an active participant in the search for the terror perpetrators.

Consequently, the two plotters became the “target” for the social media communities, and not just a headline in the media. The FBI then decided to release photos of the alternative key suspects that they had identified. It is highly probable that FBI would have held on to the photos a bit longer and not actively engaged the public in the search, if the online narrative on social media was not running so fast and furious. Preemptive release of their photos by the FBI, and due pressure by the demand for instant information in the social media world, forced the terror suspects to move earlier than they had intended, forcing them into a series of mistakes.

In addition, during the three-day lockdown of Boston, over 80,000 people turned to smart phone apps, the Internet, and any available radio listening device to follow along with the Boston Police scanner. Consequently, the most trending hash tag on Twitter was #BostonPoliceScanner. All this resulted in an unparalleled public-police collaboration, as the Police advised most Bostonians to stay indoors, the social media became the medium for the resident communities to coordinate the city-wide lockdown as police went about ‘hunting’ the terror suspects. Residential communities followed instructions of the police and also spread the word on social media. Thus what emerged was a Public Partnership for Policing, underpinned on voluntary and community ownership.

Predictive Intelligence – London Riots

For the Law enforcement agencies, social media analytics can quickly pick up intelligence on high-risk behavior. This was demonstrated during London riots of 2011. After an initial lag, the London Metropolitan Police reportedly used social media data to predict occurrence of riots in specific localities. The algorithm was based on following logic flow. Geographic clusters of mobile phones were identified on a real time basis, using location data provided by Telecom operators. The mobile concentration were indicative of a mob or a crowd assembling at specific location. The cluster was then analyzed to rule out occurrences like a traffic jam or a large party/social gathering which could also result in concentration of mobiles in an area. Thereafter, the inter communication pattern between the mobiles in the concentrated area was studied. For example, within a traffic jam the inter- communication would be very low but high in case of a mob with malicious intent where the mob-leaders were found using twitter to organize the mob.

Once such a trend was identified, the ‘sentiment analysis’ of tweets within this mobile phone cluster helped ascertain use of ‘emotionally enraged or incensed’ language, and determine the ring leaders through identification of key nodes in the communication patterns of the identified mobile phone cluster. Thus, such pattern analysis gave an early warning of potential mob violence and the real-time state of the crowd’s/mob’s state of group psychology. Counter actions in such a scenario can include jamming of mobile phones of key influencers and pre-emptive arrests of the mob leaders, and more informed redeployment of Police resources for pre-emptive incident response.

Nirbhaya Rape Protests, Delhi

The unprecedented protests and social upheaval following the Nirbhaya gang rape in New Delhi, on 16 December 2012, was triggered in a large measure due to social media. As a representative example, Sikha (name changed), 19 years, was at Jantar Mantar monument on December 25 protesting against Nirbhaya’s brutal rape when Delhi Police swooped down, rounded her up along with other agitators and took them to the Parliament Street police station. Sikha fired tweet after tweet even as she was bundled into a police van. She went on broadcasting to the world all that was happening around her. “Illegally being held here at Parliament St Police Station Delhi w/ 15 other women. Terrified, pls RT,” she tweeted. It worked. In a flash, more than 1,700 people retweeted her SOS tweet. Social media analytics indicate that the message reached over two hundred thousand people and resulted in a sympathy wave leading to even greater protestors’ crowds.

As the protests escalated across the country, water cannons, baton charges, and tear gas were quick to be deployed on the streets, especially in New Delhi. In hindsight, pre-emptive intelligence picked up from social media could have helped mitigate or prevent such a volatile outcome. Most importantly, the sentiments and opinionsbeing expressed on social media could have provided the police with insights intoemotional and psychological stress points driving the protestors – the most importantfactor that the police agencies need to know to prevent escalation of the violence and to de-escalate such a situation.

Orkash Technology

The above screenshot of ORKASH Socia Media Intelligence Platform identifies  the geographic clusters of tweets when Nirbhaya rape case protests were in progress in Delhi in December 12. Further, this technology enables detailed automated analysis of the sentiments and behavioral aspects of the tweet contents, which indicated build up of resentment and fury in the protesting crowds, giving timely indication of the transformation of some segments of the crowd into a mob, and their psychological state.

This kind of analytics and data mining of social media feeds, however, requires a complex architiecture of unstructured-data mining tools, hardware and services, (and policy controls) in the form of a Social Media Intelligence platform because of the large amounts of data to be analysed in real time. This also needs a Data Sciences approach to sentiment and behavioral analysis of the comments and traffic patterns, and temporal analysis about anticipated events. None of these are easy or readily available technologies in the current state of things!

Community Engagement for Law and Order

A recent research report has established that nearly 45% of the 100 million plus Indian web users, most of them from urban areas, connect on social media to discuss politics and social issues. Only Arab countries scored higher than India on this account. Thus, any and every Indian state agency that is a stakeholder in the Law and Order domain will need to build up expertise on analyzing social media inputs, for this is an excellent platform for listening to community and public voices. As in other countries, it is a fact that urban India resides in high density pockets. These are invariably social or ethnic clusters in large cities, underpinned further by religious/regional/linguistic/community identities.

The young population in lower-income pockets of large cities are often defined by squalor and depravity of ‘urban ghettos’, and are forced to reconcile their dreams with their economic and social reality, which often makes them susceptible to crime, drugs, radicalization and even terrorism. The RWA’s (Resident Welfare Associations) or local community leaders in such pockets are an ideal channel for the Police to tap into, and grasp, the human angle context to crime and its prevention dynamics. However, such community policing is through traditional physical interface due to perceptions of intrusion or trust. In such situations, social media accords an ideal forum for the Police to engage with the ‘Mohalla’ or local community populace without an ‘intrusive’ forward presence. This has been done fairly successfully by the Police in UK in its policing programs for vulnerable community groups.

Various police forces across the world have set up a social media monitoring facilities. These “Social Media Labs” monitor the likes of Facebook, Twitter, Google+ and other prominent social media platforms to measure changes in mass moods and track matters concerning public law and order. Police teams across the globe are also keeping a vigil on widely discussed and trending topics, in order to tie social media and criminality together. The NYPD (New York Police Department) has a program to mine social media for information about “troublesome house parties, gang showdowns and other potential mayhem”, and so has Mumbai Police recently created a Social Media lab.

Future Challenges & Opportunities

The perennial challenge for any Police department is that the amount of data covered by social media posts, updates, and tweets, will be next to impossible to monitor using traditional technology. This requires large scale infrastructure and Big Data scale of mining and analytics for textual unstructured data alongwith automated cognitive and temporal analysis. ORKASH (www.orkash.com), alongwith the likes of IBM (the Watson project), is amongst a handful of companies worldwide with the technology to do so. Of course, the inevitable dilemma surrounds the issue of privacy. Without a warrant, what information should law enforcement be able to access? Where is the line to be drawn insofar as digital intrusion is concerned? In potentially life-threatening situations, should social networking sites provide information and personal details? Though such questions may remain unanswered in the near future, the peril would be greater if they remain unasked.

Police forces have regularly received a “shot fired” message via Twitter and suspicious person reports on Facebook. Additionally, in large scale disasters scenarios, e.g. earthquakes or large terror strike, Social Media can be used for seeking and arriving at ‘situational awareness’ and optimising the incident response efforts of emergency services in the rapidly changing and confusing scenario of a disaster. More about this in our next blog post.

In conclusion, in a manner similar to beat-patrols, the Police forces will need to patrol the virtual world of social media. Be it for ‘connecting’ with the people, community-police partnerships, demonstrating presence, picking up incidental information or analyzing the conversations for pre-emptive intelligence, social media accords an unprecedented opportunity.

The accompanying snapshots illustrate the Social Media Intelligence solution developed by ORKASH Labs. They can be customized for specific Policing requirements.

Assam Roits Final

This snapshot depicts network linkages generated form ORKASH’s Social Media Intelligence platform. The graph above reveals linkages between various twitter handles commenting on the Assam riots and the User IDs (blurred out) involved in spreading rumors and   provocative  content targeted at  one particular community in cities like Bangalore and Pune, which then led to exodus of people of North-eastern origin from these cities. Font size of the handle indicates its significance in terms of influence. 


, , , , , , , , , ,

5 comments until now

  1. Vrinda Aggarwal via Facebook @ 2013-06-07 14:21

    The changing sceenrio of social interactions and relationship network with the advent of social media needs a change in the way traditional policing, law & order functions are seen and executed. This blog post provides a comprehensive view of the scope of social media in policing in India today….

  2. Col Sandeep Sudan @ 2013-06-07 16:43

    This is a very useful and a handy tool available to the Law Enforcement Agencies to deal with civil unrest situations. Today, more than half of India’s population is under the age of 25, with 65 percent of the population under 35. As a result the demograhic profile of India is such that the young generation is very tech savy and uses social media platforms to communicate within their social networks.

  3. Agree completely. The primary enabler of this capability is that sentiment anaylsis of social media posts, especially of a swelling crowd, may give an advance notice of mischevious or malicious intent.

  4. Vrinda , you would be amazed to learn that Law and Order agencies across the world are substantially increasing their vigil efforts on social media, with every passing day , on widely discussed and trending topics, in order to predict the pattern of criminality, mob mentality , unrest and mischief.

  5. Shailesh @ 2013-06-12 16:41

    Nicely captured. It remains to be seen how effectively can its potential be realised by the law enforcement agencies while avoiding biases and maintaining the accuracy levels. Political motives might also act as a roadblock towards implementing this supposedly great tool.

Add your comment now

Leave a Reply