6 Trends in Data Science for business

A digital world creates data. There are big data, small data, relevant data, and irrelevant data. In other words: we live in a world that increasingly becomes driven by data. That comes clear when suddenly, for example, due to a pandemic, systems are being disrupted. Hence, Data Science is an important field of science. What are the trends in data science for business?

Marco Brambilla, developer and teacher at the EIT Digital Professional School course Data Science for Business Innovation, who is also Professor of Data Engineering and the head of the Data Science Lab at Politecnico di Milano, shares six trends in Data Science that are impacting the way business will be done.

Data Integration is the first trend Brambilla mentions. 'When we aim at extracting the maximum value from the data, we do not rely on analysing just one source of information, just one data source at a time, but we tend to build at the original description of reality by integrating multiple resources, coming from different data providers. Today, we have the capability of capturing the data and analyzing it and transforming it into actionable value. We can use the data to build information and knowledge and improve and optimize any kind of processes and activities that we perform. This can be business processes within a company, improving quantitative indicators, improving commercial viability, brand awareness, or any kind of objective that a company can have.'

In addition to using multiple sources of data, another trend is the rise of mixing techniques for data analyses. 'To be efficient in extracting value from data, we combine multiple techniques. These techniques are, for example, machine learning, statistics, descriptive and predictive analysis. Using multiple different techniques together can help to maximise the information extraction from the data.'

Going deeper into the technical aspects, Brambilla sees deep learning and machine learning as techniques for all kinds of information. 'There is a huge amount of research going on in improving these methods. Large organisations spend research effort and time building theoretical approaches and training these approaches with a huge amount of data. That is a huge convenience because it means you can rely on any other large organisation, for instance, Google, that takes care of building and training processes for these methods. You can just reuse what has been built. That is a huge power we have today and that has revolutionised already a few sectors. Natural Language Processing – basically text analysis - is used for analysing articles, customer opinions, and social media content. Also, image analysis for object recognition, automatic description of images and so on are very convenient techniques from a business perspective, because you have an automatic detection and monitoring of things that happen and automatic product descriptions for example.'

Applications of data analysis improve or change businesses, says Brambilla. 'In the industrial sector, this could mean predictive maintenance, predictive analysis over the status of devices, infrastructures machinery, industrial installation and so on. Another sector that has changed dramatically is market analysis and customer analysis, customer profiling. Because all kinds of new data, like social media and online reviews by customers, are applied to these techniques. Also, customers now can interact better than before with automated customer support services, like chatbots. These have been around for a while, but these approaches are getting more realistic. They react when the customer talks to them. Speech recognition also has become very good. Today you can talk to your car or your phone or another device and make it do things for you. This is changing the user interaction paradigm, implying changes and improvements on the business level.'

Data Science will change the way people learn. 'The way companies and individuals are interpreting their time is changing. They perceive the need of spending more time learning.  Companies accept the fact that professional training, continuous learning for their employees, empowering them with more knowledge and understanding of trends is important. It is not about just the technical level, but also about the roles for what I would say the middle managers, project managers or decision-makers. Companies understand better now that data is becoming crucial for creating value. So, they need to empower these roles to become closer to the data-driven expertise that they need. There is a huge request for data science, data-driven approaches at companies. And especially for intermediaries between technical data science implementation teams and business managers (we can call them the business translators). Much more than in any other software computer science field.'

The EIT Digital Professional School has two professtional courses to attend:

• Data Science for Business Innovation - Italy (in collaboration with Politecnico di Milano)
• Applied Data Science Bootcamp - the Netherlands (in collaboration with Bright Cape)

For aspirant master school students EIT Digital offers a two-year master programme Data Science.

Read also the interview with Marco Brambilla: Creating Value from Data