Third year in a row, Women in Data Science Zagreb has been held around the idea of supporting women who practice data science, the ones thinking about becoming data scientists as well as educating general public on data. This year, WiDS Zagreb 2020 was a 2-day event and the conference was held on 5th March in HUB 385, Zagreb.
With the slogan “Data has a better idea!” two Croatian WiDS ambassadors – Andrea Knez Karačić & Marija Drašković – have set their goals to show how data indeed has a better idea on how to improve or even save lives! WiDS Zagreb 2020 event also had 3 technical data science related workshops held as a pre-conference day.
The conference itself had 17 speakers, 12 presentations and 1 panel discussion. The topics varied from using data in commercial to health-related issues, and below are only some of the ones presented at the conference.
Check out the full list of all presenters and thewhole event agenda on the link https://www.widszagreb.org/.
This year, we were proud sponsors of the event with two data scientists of our own who showcased how Mercury PSI handles data.
Data makes everything easier
Not being a data scientist doesn’t mean that we are not around data. The truth is that all of us are around data and we create it all the time, especially with the smart devices that became the extension of ourselves.
90% of world’s data was created in the last 2 years
And all this data can be used in many ways. As the keynote speaker Kristina Šorić from RIT Croatia told us, it is already being used in creating the next generation of digital supply chains. More precisely, data is helping them create an effective supply chain that also improves trust and transparency with their clients. It also helps create better algorithms for demand forecasting.
Sanja Seljan, from the Faculty of philosophy in Zagreb had somewhat different experience with data as she explained her work on digitalization of pharmaceutical documents, one project among many. Data helped her and her team automatize classification of numerous documents used in pharmacy, which vary from the ones written on a type-writing machine, to those with Latin phrases, and even had some damaged documents.
Data works best for trying to make sense out of something. And what better way to use data than in making sense out of unstructured texts. Ana-Marija Petrić, from Poslovna Inteligencija, worked on such projects where, among many things, she and her team made algorithms dealing with language and semantics. One such algorithm was made for summarizing articles.
Sports can also have benefits from data analysis when data is actually an image of an object. Marina Ivašić-Kos, from the Department of Informatics University in Rijeka, presented her work on automatic detection and analysis of athletic activity in handball players. Object detection and action recognition is just some of the things that go into this complex algorithm that can be used for performance analysis and ultimately preparing strategy for improving performance.
Data saves lives
Ana Dumić, a data scientist from Hrvatski Telekom, gave an account of a very personal story where collecting and analysing data saved her child who has a very rare genetic disorder. From the first day when the child was diagnosed and when the epileptic-like seizures began, Ana and her husband started collecting data on what triggers these seizures.
Collecting data and further analysing it made them realize that food was one of the main triggers and are now completely dedicated to making a unique diet along with standard therapies. With the help of collecting data, from having 10 seizures a day, currently their child is seizure-free.
IOLAP’s Barbara Pongračić also explained how collecting and analysing data can save lives but this story was about predicting earthquakes
There are more than 1 million earthquakes in a year and 8000 every day.
Even though the technology that goes into predicting earthquakes such as GNSS (Global Navigation Satellite System) is on a very high level, predicting anything including earthquakes is very hard without enough data.
Data means profit
Data monetization and big data are the buzzwords of 2020, especially in financial, banking and payment processing industry. Data can be monetized by mere generating economic benefits from available data sources or it can be the monetizing of data services.
Jelena Kolega and Darja Barišić from Mercury Processing Services International have been progressively working with data in the past years, especially when PSD2 regulation came to their data playground. Since we are in the payment processing industry, we don’t just have big amount of data to process but also we have to process it according to strict regulations that come to force, which makes data extrapolation and modelling more complex.
To make this process easier and to get to know the data better, Jelena and Darja suggest their data analysts to “go back to the sandbox”. This means experimenting on basic sets of data with the aim of better understanding that will ultimately lead to bigger expertise, clear data structures and and finally better quality of data itself.
Collecting and using data can mean higher conversion and so told us Taja Kuzmić from Five and Andrijana Brkić from Lebesgue. Andrijana focused on the mathematical process of analysing e-commerce data with the goal of reaching more conversions, while Taja spoke about using data and data modelling to track user’s behaviour and activity to make a more personalized message or design for reaching more conversions.
The video of the whole conference is available on the link below: