Can data analytics and machine learning prevent cyberbullying?

The internet is one of the most influential and important innovations of humankind. 

Since its inception in the 1990s, the web has made fundamental changes to the way we run our businesses, make payments, and even seek medical treatment.

The impact of the internet weighs heavily on our daily lives too. We depend on it for everything from consuming entertainment to getting our daily dose of news. There is nowhere better to gauge the impact of the internet than the way we communicate and interact with our social circle.

With Gen Zs growing up and relying exclusively on the internet to communicate with their peers, family, and friends and with the availability of platforms like Facebook, Twitter, Whatsapp and Instagram, our communication landscape is richer than ever.

Unfortunately, it’s not all a rosy picture. While these communication platforms are immensely important and useful, they are clouded by rampant cyberbullying.

Cyberbullying is becoming more prevalent among the younger generation on all social media platforms as well as other online forms of communication and idea-sharing like YouTube, Reddit, and Quora.

In fact, studies show that more than half of all teenagers become victims of cyberbullying, and almost as many teenagers participate in the practice. 

This issue can be resolved with the power of data analytics and artificial intelligence. Let’s explore how.

Data analytics powers improved social media policing

While cyberbullying has been reported on all internet-powered communication platforms, certain platforms are more susceptible to these kinds of incidents compared to others.

Twitter and Facebook, for example, have some of the highest rates of cyberbullying across all platforms. Specific chat groups and online communities also report a high number of incidents, and are now known as cyberbullying hotspots.

Fortunately, all these communication and opinion sharing spaces have options to report cyberbullying. The hundreds and thousands of cases reported each day allow social media companies to create a visual representation of bullying hotspots with the power of data analytics.

Once these hotspots have been identified, social media policing teams can be deployed to these spaces virtually to prevent bullying and take certain steps to block, ban and hold perpetrators accountable.

AI can help us predict cyberbullying

Studies reveal that individuals who are likely to participate in cyberbullying will show certain signs of behaviour on social media platforms through their posts and comments. The research concludes that monitoring precursors to cyberbullying can help internet companies take a more proactive approach towards preventing the practice. 

With more than three billion active users, however, monitoring online behaviour at all times is not a possibility with traditional data analytics.

The good news is that modern advancements in big data have helped create sophisticated AI speech recognition systems that can identify nuances in written speech.

These machine learning tools detect abusive and damaging online behaviour and alert authorities. This not only helps ensure the safety of likely victims but also helps rehabilitate potential cyber bullies.

AI can help us censor abusive comments, opinions and posts

Cyberbullying takes many forms and in the age of misinformation, even news articles and videos can be used to bully entire factions of the online community. 

Traditionally, social media companies have used manual content moderation to censor offensive content. Manual content moderation, however, is time-consuming and unreliable owing to human biases and errors.

Most platforms are now using algorithms powered by AI to detect and censor abusive content. This ensures potential attempts at cyberbullying can be thwarted before it causes serious harm.

Cyberbullying in all forms can be prevented with the power of data analytics and AI

In this day and age, the internet impacts every single action, including how we communicate with others. Although social media platforms allow us to communicate and stay connected with the people around us, they can be home to harmful phenomena like cyberbullying too.

Fortunately, with advancements in data analytics and artificial intelligence, we’re one step closer to making the internet a safer and healthier space for users.

Using advanced analytics and AI to improve public service

AI and advanced analytics can improve public services. At the moment, public service has problems, like long queues and delays in service provision that could stand to be a bit more efficient. However, despite the inefficiencies, there is also a wealth of opportunity, thanks to data analytics. Most government and non-profit organisations have access to large data bodies that could help improve the rate and quality of public service, provided the data is analysed properly.

This is where AI and advanced analytics can become handy. With proper application, data analysts can help transform public service to make it more efficient and responsive.

The need for advanced analytics and AI in public service

At its core, advanced analytics refers to a platform that can breakdown data with granular precision when analysing a diverse dataset. This is invaluable considering that most public service organisations have access to a rich, diverse dataset that could transform service delivery and decision-making.

For example, using advanced analytics, government health organisations can draw parallels between the chances of heart disease occurring with the language used in Twitter posts. Analysis on this level could transform the way organisations make policy decisions, allowing them to be more proactive in their decision-making.

There is also the question of AI. One of the reasons why public service tends to lag behind its private counterparts, in terms of operational efficiency, is because of menial labour tasks. So many menial tasks are still in the hands of employees, which holds up the rate of service delivery.

However, thanks to AI, these menial tasks can now be conducted by machines at a faster rate. This will help improve the rate of service in public organisations because employees no longer have to waste a significant portion of their time performing trivial tasks. While the AI handles more monotonous tasks, public servants can do more advanced work at the same time, improving productivity and service delivery

Furthermore, there is also the degree of accuracy to consider. It is easy for government employees to make mistakes, considering their workload. Advanced analytics and AI can help improve accuracy by a significant margin. Better accuracy means fewer reworks required, which means more productivity (which means fewer headaches for all involved!). Through proper use of advanced analytics and AI, we have the opportunity to improve the quality and efficiency of public service.

Preparing public services for advanced analytics

While there is no denying the benefits of AI and advanced analytics in public services, it does entail a degree of forethought and preparation.

This is because advanced analytics and AI is not as simple as installing software and running it. It needs to be installed and administered correctly by the right team. Even after systems are installed and working, there needs to be a workforce skilled enough to take advantage of the data analytics systems.

Organisations need a force of IT personnel that can extract data from the analytics platforms and generate useful insights from it. Even beyond the core analytics personnel, managers (or anyone in the middle management level) need to be skilled enough to identify opportunities presented by data analytics platforms.

Furthermore, public servants will need to adapt to working with machine-assisted, decision-making algorithms, which not only leads to a technical shift in how the organisation works but also a cultural one. This is because everyone’s role within the organisation will shift, along with their skillset. In some cases, there might even be a question of whether they need to build in-house analytics expertise or work with a consultant, which could have further implications, in terms of contractual obligations.

Working towards better public service

To improve public service, we need to explore all potential avenues, including the use of advanced analytics and AI. Advanced data analytics can help organisations move more efficiently than before. Furthermore, AI can help augment the rate of productivity because of several factors, like the ability to automate several functions and even eliminate human error from certain processes.

Given all these factors, advanced analytics and AI can help organisations by a significant margin. However, to make full use of these analytics platforms requires a complete shift in the way the organisation works, from both a technical and cultural perspective.

How AI impacts Business Intelligence Applications

The business intelligence applications industry is growing each year with research reports from organisations like Forex Tribune projecting significant growth. There are several reasons for this growth, the proliferation of big data, the expansion of IoT and the need to better understand the collected information are all contributing factors in the growth of BI. However, despite the high value of its contributions, BI can still come with a few improvements. Improvements made thanks to Artificial Intelligence (AI). Hence, I will explain what BI is, how AI improves technology and what it means to organisations.

Business Intelligence – What does it entail?

For those not in the know, business intelligence (BI) refers to systems designed to streamline the collection, processing and analysis of big data. Organisations can make sense of the large volumes of data they collect to sharpen their business processes and make smart decisions. As useful as BI is, there are some shortcomings that hamper the value it contributes to businesses. For example, there is a limit to BI capacity and the sheer volume of data collected is pushing that limit. BI will benefit tremendously from the application of AI because the latter augments the former.

How AI boosts business intelligence applications

AI expands BI functionality

Artificial intelligence boosts the functionality of business intelligence applications. With AI, business intelligence is better placed to breakdown large volumes of big data into granular insights, allowing organisations to better understand the value of smaller components into a larger picture.

Then, there is the issue of real-time insights, given that in their current iteration, BI can process and visualise big data, but cannot predict trends or generate real-time insights. However, AI incorporates the latest technology, like machine learning to deliver real-time insights about trends that will take place in the future. Thus, expanding the functionality of business intelligence applications and improving its value to organisations.

Close the gap

AI-enabled BI allows businesses to develop critical insights into data not examined before. Business intelligence applications powered by AI can examine fresh data and identify any trends relevant to the organisation. AI also allows BI to utilise the latest technology, like predictive analytics, machine learning and natural language processing to expand insights presented. Organisations are no longer satisfied with a visual dashboard of big data trends, they need tools that can close the gap between visual representation and actionable insights. AI-powered BI can help close this gap.

Simplifying a complex process

Surveying big data is often a complex process even with business intelligence applications. Professional data analysts have to survey 100s of charts and dashboards to get the insights needed. However, AI technology can simplify the process. The simplification happens because AI-based technologies, like NLP and machine learning, are closing the gap between machine and human communication. AI technologies allow machines to better understand human language and vice versa, making it easier for data analysts to find connections and insight. AI-powered BI allows organisations to tackle a much larger variety of data, including structured and unstructured data to get more detailed, comprehensive insights.

Solve problems related to talent storage

Business intelligence presents data findings on a visual dashboard, however, with data coming in from multiple sources, it becomes harder for dashboards to present the data in an easy to read format. However, with AI, the information can be defined at scale, making it easier to gain actionable insights. There is also the issue of talent. As of 2019, there is a severe shortage of data analysts. The right processing software can help alleviate some of the problems from a talent shortage by performing some of the functions normally delegated to a data analyst.

AI is the future of business intelligence applications

Business intelligence applications have been a tremendous asset to organisations, allowing them to better understand big data. However, times change and the needs of organisations evolve. They not only need intelligence software that can process and present data into visual findings, but they also require software that can predict trends, anticipate actionable insights in real-time and process a variety of data. This is where AI comes in since artificial intelligence allows organisations to breakdown big data into granular levels, so it is easier for organisations to make smarter decisions.

AI and data analytics – What is the connection

AI and data analytics have been used in conjunction for some time, to the point where people rarely distinguish between the two terms. However, as data analytics and AI capabilities become more widespread and applied to different business operations like marketing and supply, it’s essential to understand the difference between the two and what role they will play in business operations. That is my intention for this blog post: To explain the connection between analytics and AI.

How is AI and data analytics defined?

Before explaining the connection between AI and data analytics, we need to take a moment to define the terms. AI or Artificial Intelligence is technology designed to emulate the human mind, particularly in areas such as analysis and learning. AI is designed to draw conclusions on data, understand concepts, become self-learning and even interact with humans.

Data analytics refers to technologies that study data and draw patterns. However, the function can vary based on the type of technology. For example, descriptive analytics can study data to describe what is happening, while predictive analytics can predict what will happen based on current occurrences. Furthermore, when it comes to data analytics, it is not a single product. It is a rich ecosystem of programs ranging from the basics like descriptive analytics and BI to more advanced programs such as data mining, forecasting and pattern matching.

The connection between AI and data analytics

AI and data analytics are connected because the former boosts the capabilities of the latter to deliver deeper and better insights beyond what human analysts can do. The point is best demonstrated through an example. A supermarket chain offers a loyalty credit card program that allows customers to accumulate points when they use the credit card. The points can be spent on a day at the golf course.

Descriptive analytics reveals that, out of 10,000 members, over 1,500 cashed in their points for a day at the golf course and all of them were middle-aged men. Predictive analytics reveals that with a 10% increase in advertising, the supermarket chain will see a 20% increases in the number of middle-aged men cashing in on their reward points.

However, with AI, the bank will discover the number of members who live near a golf course, those who like the sport but don’t go, if any female members are interested in golf and set parameters for golf season in certain climate zones. Using this information, the bank can then develop specific, targeted programs for single men, single women, married couples, families and more. Therefore, AI enhances the capabilities of data analytics tools to deliver granular, micro-targeted insights that were not possible before.

Besides enhancing analytics capabilities, AI also improves the data analysis process. Since data comes from structured and unstructured sources, it must be cleaned and organised before it’s ready for analysis. Data analysts spend 80% of their time cleaning and organising data. AI can be used to accelerate this process, thus saving data analysts time and making the process more efficient.

AI and machine learning can also enhance the capabilities of data analytics models beyond their current capabilities. An excellent example is fraud prevention in insurance or banking. With machine learning, analytics models can identify fraudulent transactions in real time. The analytics models can identify fraud as it happens because data analysts feed data on past fraud incidents to the analytics models. The AI can study data, learn the patterns that make up a fraudulent transaction to identify future fraud transactions. If fraudsters change their methods of attack, machine learning can adapt by picking up on these new methods.

Key takeaways

AI and data analytics are often used together because the former boosts the functionalities of the latter. With AI, analytics technology can conduct more in-depth analysis paving the way for micro-targeted insights that are not easily found by human analysts. Complex analysis with several variables can be done quickly and efficiently with AI.

AI in data analytics also makes it easier to clean data – a vital step in the analysis process. It’s important to understand that AI and analytics are not the same and should not be considered as such because AI is part of the analytics ecosystem. Companies must understand the difference and be willing to use the technology if they wish to gain an edge over their competitors.

Want to learn more about AI, machine learning and data analytics? Our blog has all the information you need.

AI has revolutionised human-machine interaction and has played a vital role in data analytics. SAS provides cutting-edge AI solutions that can improve your organisation’s productivity, and using our Selerity analytics desktops can give you access to this technology. Consult our team today to learn more about AI solutions in SAS.

The benefits of SAS Viya’s machine learning and AI capabilities

Since its announcement last April at the 2017 SAS Global Forum, SAS Viya has been heralded as the analytics platform’s stepping-stone into the future. As a cloud-enabled platform, with a powerful in-memory analytics engine that enables quick, accurate, and consistent results at all times, Viya is scalable and has the processing power required to address some of today’s most complex analytical challenges. By combining all its existing functionality with machine-learning and artificial intelligence technology, SAS software is not only becoming more efficient, but smart in the process. This scalability and dexterity are precisely what makes users of SAS software optimistic about the platform’s value in the future.

So, how exactly does SAS utilise machine-learning and AI capabilities? What noticeable benefits has this technology fostered? We explore all that and more in the following sections.

Which SAS Viya product has the strongest machine learning and artificial intelligence capabilities?

While SAS Viya comprises of at least 12 varying products – from SAS Data Preparation to SAS Visual Text Analytics – SAS Visual Data Mining and Machine Learning is by far one of the platform’s most intelligent and insightful. Given that the solution runs on SAS Viya, it is bolstered by the platform’s ability to manage any analytics challenge. Its relative user-friendliness means that anyone from data scientists to business analysts to developers, and executives can collaborate with another to realize insights and results faster.

SAS Visual Data Mining and Machine Learning comes equipped with an incredibly broad set of modern machine learning, deep learning, and text analytics algorithms that are all accessible within a single environment. This makes the solution ideal for all kinds of business users, given the solution’s diverse analytics capabilities that include clustering, different modes of regression, random forests, gradient boosting models, support vector machines, natural language processing, and topic detection – to name a few. SAS users not only gain access to a platform that is highly functional, but one that is equipped with powerful predictive and decision-making capabilities that were previously limited.

Now, let’s look at how this benefits end-users – we begin with how the platform helps solve complex analytical problems much faster.

• Complex analytical problems solved much faster

Since the software runs on the latest edition of the SAS Platform – SAS Viya – it has the ability to deliver predictive modeling and machine-learning capabilities at unprecedented speeds via powerful in-memory processing. Given the processing prowess and persistence associated with this in-memory data, the need to load data multiple times during various iterations of analysis is no longer required. This means that multi-user collaboration has been simplified where users across all segments of the business and/or organisation can explore the exact same raw data and build their respective models simultaneously. Through the SAS Viya platform, analytical modeling can be done in a matter of minutes, which enables organisations to find answers to their questions and challenges much quicker and efficiently.

• Greater simplicity and improved usability

A few years ago, the idea of building complex analytics models to drive data-based decision-making within an organisation was heralded as an extremely difficult task that required significant technical expertise. However, through the evolution of the SAS Platform and the development of SAS Viya, users now have access to a platform that comprises of interactive visual and programming interfaces that significantly reduce the amount of time it takes to set up data, build complex and insightful machine learning models, and, finally, make decisions based on these insights.

Even users who lack coding expertise and knowledge can leverage the platform by generating advanced machine learning algorithms via the platform’s built-in visual drag-and-drop interface without ever having to know or drop a line of code into the system. This is something we like to call complete organisational empowerment.

For users who are more technically-versed, data sources and code snippets can be shared among themselves and across departments to better improve organisational collaboration. Additionally, business users of the platform would not have to exclusively know how to code in SAS – other languages, such as Python, R, Java, and Lua can be used to code as well. The SAS code is automatically generated behind the scenes!

• Leverage machine learning to make the right decision after exploring multiple options

With machine learning capabilities built in, business users have the benefit of leveraging the platform to evaluate and compare all available options/approaches, prior to making or recommending a decision. Scenario-based decision-making is facilitated via the system’s “automated model tuning”, which lets users identify the best-performing model.

The machine learning programs can integrate both structured and unstructured data, enabling users to derive more insights from new data types by adjusting their models accordingly.

SAS Viya has many benefits, with its machine learning and artificial intelligence capabilities being among its best features. If you would like to know more about Viya and its features, in addition to how you can install, administer, and host your organisation’s own SAS environment, feel free to reach out to us, or stay tuned to this feed.