How to tell when negative sentiment becomes a threat to your business

There are some 500 million tweets and over 4 million new blogs posted every single day. Each of these sparks another conversation which could house potential threats against an organisation. And we haven’t yet mentioned Facebook, Instagram, Reddit, Flickr, Medium or any of the other dozens of social sites and forums where people post online. And if you thought that was a lot of noise you have to remember the dark web too, where many cybercriminals go to engage in nefarious activities with the protection of a Tor browsers anonymity. 

The point here is that the internet is full of noise. Monitoring all of that and then cutting through the noise to detect relevant potential threats requires the right tools. 

What is Sentiment Analysis?

Sentiment analysis, in short, is analysing the language in online posts and comments to determine the underlying emotion behind what has or is being posted by an individual or group.

Determining online sentiment doesn't just allow you to understand better how your brand is performing and how people feel about your business. It can also be used to manage crises and spot potential threats to assets or staff.

Without sentiment analysis, data can be misleading. Sentiment gives data extra context which allows it to be better understood enabling a more effective and accurate response to the potential rtisks. 

It also allows you to differentiate between when a negative comment is simply that, a negative comment, or when it needs more serious attention because, for example, it’s evolving into a physical threat.

Where and How do we Measure Sentiment?

Any text that gets highlighted by Signal OSINT software can be run through our sentiment analysis tool, Spotlight. This allows users to reduce the amount of noise and focus on the threats.

Sentiment can be expressed anywhere online, this might be through social media, in the comments of a blog or even in a dark web forum. Signal allows you to gather data from a huge array of open intelligence sources including (but not limited to) social media and dark web forums.

How can Sentiment Analysis Be Used for your Business?

Emerging Threats

Sentiment analysis can be an incredibly useful tool for those that wish to identify potential risks which might evolve into tangible reputational or physical threats against, employees, executives, brand or assets.

Managing Reputation

Your brand’s health and reputation are important. Having a tool that allows you to analyse the overall sentiment towards your brand and associated keywords gives organisations a bigger and better overall picture of their brand which can be a game-changer for launches of major events or analysing the success of a large marketing campaign.

Evolving Crises

When it comes to dealing with current and evolving crises having up to date and detailed situational awareness, gained through an OSINT tool such as Signal can make a huge difference. However, as we have mentioned before, there is a huge amount of noise out there. So, how do you determine which comments, which posts are relevant and need monitoring?

The answer is to use Signal to create specific filters and then run identified posts through our sentiment analysis tool “Spotlight”. This allows users to both quickly identify emerging threats and to then stay on top of these risks as they are evolving in real time.

Moving Your Marketing Forward

Social sentiment is a powerful tool for understanding the relationship between your brand, your customers, and your competitors. If you measure it regularly and act on what you learn, your team can create targeted marketing strategies to keep up with the ever-changing demands and opinions of your customers.

How do you determine when Negative Sentiment Becomes a Threat?

One of the key methods used by our software and our analysis team to tell whether or not a comment is a threat that needs more attention is the repetition of negative sentiment online by an individual or group.

For example:

• Does a particular author of a comment or post have a long history of bad-mouthing an organisation or expressing negative sentiment?

• Have they repeated the same negativity on multiple sources?

Even if they aren’t directly threatening any physical or tangible action against the organisation, if there’s enough online commentary from a single individual or group then this could escalate and it may be smart to further monitor.

You can then set up a search using our filters to target this individual or group so that you don’t miss if this negative sentiment becomes a physical or reputational threat. 

Secondly, using Spotlight, users can identify posts expressing dangerous emotions such as anger, or disappointment. Both if repeated enough should be addressed. Posts expressing anger are likely to indicate a physical threat and should be monitored for that, whilst the posts expressing disappointment may hold reputational risks. 

Summary

Sentiment analysis tools like Signal’s Spotlight can help security teams form a broader and more detailed overview of the situation to better understand the potential and emerging threats. It allows them to target their online searches and cut through the noise to identify key threats. All of this essentially means a more efficient and more effective security team.

You also might like:

Critical Security Intelligence for the Financial Services Sector

Critical Security Intelligence for the Financial Services Sector

A Financial Services Use Case

Since its conception Signal is proud to have developed a strong use case across the financial sector. Proving to be an invaluable tool for a number of financial services organisations including several multinational and Fortune 1000 companies. 

Signal open source threat intelligence platform now forms an integral part of their security teams’ toolbox, assisting across concerns, from detecting and managing physical threats to assets, reputation or VIPs, to aiding cybersecurity teams with efficient methods for open and dark web monitoring.

Physical Security

Using Signal, insurance providers, financial companies, and banking organisations can not only gain an overview of emerging threats in real-time but also target key areas and assets that they want to closely monitor creating a hyper-relevant stream of actionable real-time data.

For example, banking organisations use Signal to monitor geo-located information online, focusing their web monitoring on key locations of particular interest such as ATMs, head offices, or VIP locations. They can additionally run the focused data that surfaces through our sentiment analysis software. This allows them to help cut through the noise and quickly identify online chatter expressing negative sentiment so that they can more efficiently distinguish threats. 

A secondary use of Signal for the financial services sector is monitoring their competitors’ security threats which may also be affecting them - even if they don’t know it yet. This allows their security teams to predict potential threats even before they emerge.

How Signal has been used in the Banking Sector to Detect and Deter Threats

In 2019, one of Signals clients was able to prevent a particularly worrying case of attempted fraud. An employee at that bank was being harassed online in an attempt to instigate said employee to provide confidential information of customers to hackers for use in fraudulent activity.

Thankfully, Signal was able to spot this before matters progressed, preventing potentially hundreds of thousands of dollars worth of damages.

Signal has also been used to pinpoint worrying sentiment against some of our customer’s executive teams.

Cyber Security Threats

Signal is also used by customers for cybersecurity intelligence to determine risks such as:

• Fraudulent statement packs; 

• Fake bank statements;

• Phishing websites (using our in-built image recognition);

• Credit card and account credentials for sale on the dark web.

Security teams use Signal to monitor both open and dark web conversations. This enables them to keep abreast of new and developing global trends and methods used in cyber fraud which is effecting their industry as a whole. Using this data they can actively develop strategies to prevent and tackle new and rising methods of cybercrime. 

Signal also allows these teams to monitor for data breaches by alerting security teams as soon as chat is identified around potential data breaches which could affect the organisation itself, or their customers.

Ways Signal Helps Financial Services

Identify negative sentiment and potential threats to customer assets.

Detect and intercept stolen credit card and account credentials for sale on the dark web.

Monitor targeted locations to create a focused stream of actionable real time data.

Signal allows our customers to analyse emerging global trends, detect threats in real-time and then form appropriate security strategies to counter these potential threats as or even before they fully reveal themselves.

For the financial services having this targeted and focused stream of accurate and relevant data is vital to ensuring the safety and security of their customers, assets, and employees.

The Power of Emotional Analysis - introducing Signal Spotlight

Signal Spotlight provides a real-time overview of the emotional state of Signal search results. Using Signal Spotlight, Signal users can better understand the prevalence and drivers of emotions and what is happening in real-time.

Spotlight taps into the results from Signal search criteria across many data sources to better understand the prevalence and drivers of emotions. For example, during an incident or event, an important attribute is how people are feeling about what has happened and how the emotional state may be changing real-time as that incident/event unfolds.

The Spotlight underlying technology uses a large vocabulary of emotion terms that were compiled from multiple sources, including the ANEW and LIWC corpora, and a list of moods from LiveJournal.  In addition, a crowdsourcing task was run to organise these terms against Parrott's hierarchy of emotions. The emotions are colour-coded using a dataset of affective norms provided by the Center for Reading Research at Ghent University.

Spotlight leverages technology and research undertaken by the Language and Social Computing team in the Digital Economy Programme of CSIRO's Digital Productivity Flagship and originally developed as a joint project between computer scientists at CSIRO and mental health researchers at The Black Dog Institute.

Reference

Milne, D., Paris, C., Christensen, H., Batterham, P. and O'Dea, B. (2015) We Feel: Taking the emotional pulse of the world. In the Proceedings of the 19th Triennial Congress of the International Ergonomics Association (IEA 2015), Melbourne, Victoria, Australia, August 2015.

Case Study: How Emergency Management Victoria Leverage Social Media Intelligence To Their Advantage

Justin Kibell, Manager of Operational Intelligence, Emergency Management Victoria

We spoke with one of our customers, Justin Kibell - the Manager of Operational Intelligence for Emergency Management Victoria (EMV), to see what he thought the most important uses for monitoring online data were, when emergencies arose.

Here's what Justin had to say:

One of the key principles we use in our Intelligence section is to consider a broad range of information from different sources and agencies. Open source is a key source of information directly from the community, but it comes with a lot of noise.

It can often be difficult to locate the key pieces of information and this is why we have been using Signal to assist us with efficiently gathering and collating open source content across a range of platforms.

The key drivers for information gathering from social media platforms are similar to that of other information channels monitored by Intelligence Analysts.  Open source information such as social media provides an opportunity to corroborate information from other sources and in some cases discover additional (important) information shared directly by the public who are on scene or potentially impacted.

"... Intelligence Analysts utilise the powerful search and monitoring features of Signal..."

Our Intelligence Analysts utilise the powerful search and monitoring features of Signal to search through various open source streams to locate information across a range of intelligence requirements such as:

1. Gauging Public Sentiment – to assist the social media team in our Public Information team, we use Signal to help determine what the community is saying about the emergency and the Governments response, both positive and negative.

2. Monitor News Coverage – searching local and interstate news websites, journalists and associated influencers, we use Signal to help obtain a bird’s eye view of what media are reporting which we pass onto the Public Information team to address and minimise misinformation.

3. Situational Awareness – using Signal to search and collate information from videos, images and descriptions of damage posted online is incredibly useful to our analysts both in building a more complete picture of what is going on (including known and potential impacts), but also in assisting our predictive services team with on scene observations which help validate their modelling such as photos with smoke plumes and fire behaviour showing.

4. Survey Damaged Areas – pinpointing the worst hit areas with observations directly from the community assists us in more efficiently deploying resources to areas with the most need. The geo-located content he at map feature quickly highlight key areas of interest.

"... using Signal for over two years now"

At the State Control Centre we have been using Signal for over two years now. Our Intelligence Analysts have positively commented on improvements to the usability and features of the tool and look forward to using the new information offered through the new “Spotlight” functionality.

Over the last three years our social media emergency management intelligence capability has grown and throughout this period Signal has been an important and valuable tool in the tool chest used by our team.

Justin Kibell,  Manager of Operational Intelligence, EMV

---

Justin's experience with Signal demonstrates the various benefits social media offers for emergency management intelligence. 

How are you using open source intelligence in your emergency response efforts?