e-Newsletter


Artificial Intelligence: More Than Meets the Eye

Artificial intelligence is beyond a doubt one of the most, if not the most, exciting areas of computer science . The Slovenian Jozef Stefan Institute (IJS) plays an important role in this field of academic research with far reaching practical applications.

The AI department of the institute led by dr Dunja Mladenić remains unknown to the general public even in Slovenia, yet its data mining solutions were used by global media giants like The New York Times and Bloomberg. Take the IJS's latest product Event Registry and its spin off company Quintelligence.  A unique tool for real-time collection and analysis of news published around the globe, the system analyzes news coming from over 100.000 sources in different languages. Sources also include blogs and tweets. A user of Event Registry can track how news spreads around the world, the development of the story and the context in which it appears. The system allows for exploring of trends as well as making predictions on how news would spread. It is an invaluable tool for media publishers and companies alike. For now the system can only analyze text based sources, but researchers are planning to develop an extension which could follow news on television.

Developed with the help of European funding, Event registry is an entirely Slovenian product based on almost two decades of research. A Few years ago the same group of researchers developed a system they called the NYT Miner – a data mining solution developed to follow the behavior of readers of the New York Times website. Other areas of the IJS's AI department include machine learning, semantic technologies, social network analysis and natural language processing and technologies. The department collaborates with a number of international academic institutions and companies. One of the companies closely linked with the IJS's AI laboratory is US company Cycorp, “a leading provider of semantic technologies that bring a new level of intelligence and common sense reasoning to a wide variety of software applications.” Cycorp is developing a project named Cyc – the ambitious project, the ultimate goal of which is to enable AI applications to perform human-like reasoning, was started back in 1984 by Texan computer scientist Doug Lenat. Cycorp’s only company outside of Texas is located in Ljubljana and has close links to the IJS. Cycorp Europe is led by Michael Witbrock, who is also Cycorp’s vice president of research. Witbrock is also chief technology officer of another IJS spin off, a company called Envigence. Based in Slokan, near the Italian border, Envigence develops environmental intelligence systems – to be used in smart communities, for example.

While key personnel from Cycorp are active in Slovenia, local computer scientists are working in leading global research institutions. Jure Leskovec, an assistant professor of computer science at Stanford University, is one of the leading experts on mining and modeling of large social and information networks like Facebook or Twitter. In simple terms, Leskovec uses huge datasets generated by social networks as raw material to analyze and identify fundamental patterns of human behavior. His models can, for example, predict which news could become viral. Not long ago Leskovec and his colleague Julian McAuley developed an algorithm that automatically divides Facebook friends into different social circles. In 2015 he co-authored a research paper on antisocial behavior in online discussion communities. A group of researchers including Leskovec developed a model which could predict patterns and identify online troublemakers (so called trolls) in advance.

At 34 years of age Leskovec is the youngest to bear the prestigious title of Slovenian Ambassador of Science. On the web he is also known as “Mr Context”. In his words, he uses terabytes of data as a telescope or “a sensor into human life”: “what wasn’t visible before is visible now”.

You might be also interested in:

GET IN TOUCH

We are here to help you.