5 Ways to Utilize Data in the Cyber Space

While people behind the IoT want to connect every item to the internet, some are concerned even with the current state. Our private and financial info is already out there for anyone to see. Then, there are all those online threats, like cyber attacks, lack of privacy, and cybercrime.

How does one fare against those odds?

Not great…

That is unless you can learn to utilize data in cyber space to elevate your security to the next level. Here are five ideas on how you can do this.


1.   User behavior analytics

We all know the Sci-Fi tropes where two people’s minds replace bodies, and no one can figure it out until the near end of the movie. The truth usually reveals itself through one person displaying strange behavior patterns, gestures, or even phrases they’ve never used before.

One can use a similar principle to detect an account takeover. Still, to teach the algorithm to recognize this, we need to feed the system with incredible data and rely on machine learning (ML) to draw all the right conclusions.

You see, different people have different types of behavior in the digital world. Acting out of character could be enough to raise some red flags.

  • Keystroke dynamics: We all type at approximately the same speed (when not trying extra hard). This pace or pattern difference can reveal another person behind the account.
  • Mouse movement: This habit is deeply rooted in our behavior patterns, and it’s nearly impossible to change it. We are clueless that we do the same thing repeatedly, so how can we change something we’re unaware of?
  • Scroll patterns: The way we scroll, slow down, scroll up, and more are also quite consistent. Same as with the above-listed two items, we’re unaware of this.
  • Browsing patterns: Everything from an average website retention time to the number of clicks we make on each page is fairly consistent over our browsing history.
  • Navigation behavior: Details like whether you go back to the home page or use the navigation screen can also indicate that it might not be you.

The majority of modern anti-counterfeit solutions revolve around these user behavior metrics.

The thing is that the only way for any of these data to be reliable a program would have to analyze incredible quantities of data. Fortunately, this data is available, and these analytical tools are up to the task.

2.   Preventing online fraud

During your research on how to prevent online fraud, you’ve likely encountered many horrifying statistics about the number of people who experience online fraud. There is a huge number of people’s information and finances unprotected from malicious online parties.

As an individual, you can be more careful who you share your info with, but organizations can utilize vast data for this purpose.

For instance, you can notice strange transaction patterns through historical transaction analysis. Let’s say someone places numerous orders, the largest of which is $10, and then, one day, they decide to spend $4,000 on your site. Wouldn’t that be a bit odd? Of course, it’s not indicative; everyone has the right to increase their spending (without providing any explanation or justification). However, you can’t deny that it’s a bit odd.

The geolocation of a card is also important, especially with the transactions’ history. If the location doesn’t match a cardholder’s address or a card belonging to a region renowned for a high rate of credit card fraud, the risk is somewhat higher.

While you don’t immediately know it’s a fraud in question, combining several factors allows you to get an arbitrary assessment (in the form of transaction risk scoring). This way, you’ll get an approximate idea of what risk you expose yourself to.

Most importantly, financial organizations can use data from previous credit card frauds to ensure this doesn’t happen again. This will allow people in e-commerce and finance, in general, to become proactive.

3.   More sophisticated threat detection

The earlier we recognize the threat, the higher our chances of protecting against them. This is especially true in the age where a malicious party with a VPN and access to digital assets (like crypto or NFT) can just vanish, never to be found again.

As mentioned in the previous section, we need to become proactive. This will allow us to blacklist malicious users and, for the first time, allow anti-malware and antivirus software to develop as quickly as malware and viruses.

There are many ways to enhance the existing threat detection methods, but what about the future? What about the evolution of these systems? There are several methods worth considering.

  • Log analysis: This method tracks all the previous activity of an account (or interactions between these accounts and your online presence). This is the key to noticing some of these strange online behavior patterns.
  • Endpoint data analysis: Data gathered by antivirus or host-based intrusion detection systems (HIDS) is invaluable. You see, this is data that are, beyond doubt, troublesome. ML can use it as a reference point.
  • Intrusion detection system: You need to know how to recognize the malicious activity, and the best way to do so is to compare it to other malicious activity. Comparing two instances is crude and inaccurate, but comparing dozens of these instances can become quite reliable. We would need incredible quantities of data at our disposal.

Remember that it’s not just about the cost of a potential breach. It’s also about the repercussions in terms of your reputation. This is one of the most important assets that you have in the digital world. Recovering after a reputational loss is never a sure thing.

4.   Security Metrics, KPIs, and Reporting

Even with structured data, measuring security is never an easy job. How can you evaluate a threat? Previously, we discussed various arbitrary numbers, but there’s still no convention on these things, and it’s still incredibly difficult to get an adequate read on various threat levels.

However, what if we could agree on some security metrics like:

  • Vulnerability metric
  • Compliance metric
  • Risk metric
  • Security awareness metric

You also need to understand the performance of your security measures or system. For this, you would need KPIs like:

  • MTTD: Mean time to detect
  • MTTR: Mean time to respond
  • User awareness index
  • Return on security investment

These metrics will tell you if you’re on the right path or if there’s something that you have to improve on. You also see if you’re getting your money’s worth from these investments.

During uncertain times, even a small security breach can be enough to ruin your company’s reputation. This is why improving these metrics helps you become more resilient.

Finally, establish a reliable reporting system to ensure that your system is ironclad. First, you need to set some ground rules, like how often this report is needed and what you should put on this report. Then, you must add some contextual requirements (point out the data you need every time). Lastly, you need to know who the report is for. This will set the terminology used.

5.   Security awareness and training

Without data, there’s no training program.

Most cyber breaches happen because of the lack of awareness of your team. So, you need to enhance their training to prevent this from happening, but you usually don’t have to start from the bottom. This means that your first step needs to be establishing the baseline. This way, you’ll develop an understanding of their current knowledge level. This will help you devise a plan/curriculum.

The best way to conduct training is through theoretical background coupled with some simulations. Ideally, you want to create a scenario of real phishing (a phishing simulation). To make this as effective as possible, you should extract all the available phishing data and figure out which scenario they’re most likely to encounter in their work. Even hiring a white hat hacker might be a great idea.

Next, you want to gamify this curriculum to ensure maximum immersion and focus by your team. This will help you send the message across far more effectively than if you used old-school teaching methods.

Training effectiveness evaluation is one of the things that you should never skip. Just because something sounds like a good idea doesn’t mean it is.

Finally, every training process can and should be improved over time. So, use all this training data and feedback to see which parts of the process performed admirably and which failed miserably. Continuous improvement is the name of the game.

With the right data utilization cyber space can become a much safer place

With so much crime out there, it all seemed like an uphill battle. Now, we may finally have a fighting chance with big data, ML, AI, and other tools on our side. Just remember that this is a never-ending process and that, before this is done, both parties have a long way to go and a lot of tools to improve. One thing is certain, it all hangs on data.