Skip to Content
Author's profile photo Former Member

Making social good from social media – unstructured data in policing

The data generated within social media is for many, a by-product of this instantaneous or asynchronous form of digital communication. But for others such as public security authorities, the mountains of unstructured information residing within social media  provides a rich source of intelligence that can be used to help keep our communities safe.

Traditionally we’ve been able to make decisions based on structured data from excel spreadsheets and graphs. This might have been extracted from a variety of sources such as forms and telephone surveys. However, life doesn’t always fit into neat little boxes and digital disruption has brought new expectations of how data is gleaned and how it is analysed.

Big Data is a buzz word that essentially covers all of the technical information that is being created by individuals, companies and organisations around the world. The volume is massive and it breaks down into two basic types – structured and unstructured data.

Structured data is any data that can be easily sorted and managed. Think spreadsheets and databases as an easy way to understand it. While this type of data is the easiest to manage, it also represents a smaller share of the total data available – about 20%.

Unstructured data (the other 80%) represents everything else you can think of. From our facebook posts to our tweets, blogs and videos that are published across the Internet, this human generated content is being amassing at breakneck speed.

Although unstructured data is a massive and almost overwhelming concept, there are ways to monitor and analyse it for public value. An interesting discussion at a recent SAP Institute for Digital Government  security roundtable centred on policing agencies monitoring social media. In recent years police have tackled the behemoth of unstructured data (in particular social media) in a constructive way to make communities safer.

Recently the Victorian Police showed their prowess in this field after discovering an open Facebook invite to an Australia Day party (Jan 26th) on St Kilda beach. With 6,000 people registered as ‘going’ and a further 19,000 as ‘interested’, the party had the potential to get out of control and contribute to the spike in drunkenness and assaults which usually take place on Australia Day.

Police have a duty to ensure public safety and drunken disorderly crowds gathering for “public parties” have a history of getting out of control and causing chaos, leading to public damage, arrests and injuries. The usual process in Australia, (as similar to many other countries) is for large scale gatherings to be registered with a permit to use the public space. Before the age of social media it’s unlikely the police would have known of this party before it happened. However, neither would have the 25,000 people indicating they may be attending.

So, you may ask, how did the police respond to this? Well, they messaged the organisers and invited themselves via the Facebook event page.


This clever and witty interaction led to the police and party organisers working together to plan a safer and more organised party for next year in a legal venue – an outcome favourable for all involved.

The police may have been lucky to have seen this post by simply having officers monitoring social media or they may have been tipped off by a member of the public. Today’s police forces have specialists who analyse criminal patterns of behaviour, including online activity but with so much activity going on in social media in real-time, the specialists need automated systems to analyse this unstructured data to highlight areas for investigation.  Analysing unstructured data in an automated manner can help policing agencies put officers in a position to more quickly and effectively prevent — or at least respond to — criminal activities and antisocial behaviour. As social media is intended to share personal stories, some people take it further than a party and literally boast or hint of their crimes! Location services can also determine a criminal’s movements and may give real time data and evidence critical for police investigations.  But unless someone is watching then who is really listening and this is where automation comes to the fore.

There are many positive ways in which unstructured data can be utilised for public security reasons. Although the majority of data is unstructured and it’s growing rapidly every second, there is still public value and capacity for policing agencies to assign resources to digital analysis to help solve crimes.

For every criminal activity solved through data analytics there will always be many more that aren’t. The rapid growth of this data and the ability to derive value from it, is undoubtedly a challenge for policing and other industries, but one well worth taking to make social good from social media.

While digital technology is responsible for enabling the creation of unstructured data, digital technology is providing us with the tools to automate the analysis and interpretation of this data.

Blog post written by Belinda McKeon and Kylie Watson from the SAP Institute for Digital Government.

To find out more about the SAP Institute for Digital Government visit, follow us on Twitter @SAPSIDG and email us at

Screenshot 2015-03-26 09.47.54 (2).png

Assigned Tags

      Be the first to leave a comment
      You must be Logged on to comment or reply to a post.