Erik Novak

Jožef Stefan Institute | IJS · Department for Artificial Intelligence

Modern cross-lingual document retrieval models are capable of finding documents relevant to the query. However, they do not have the capabilities for explaining why the document is relevant. This paper proposes a novel learning-to-rank model named LM-EMD that uses the multilingual BERT language model and Earth Mover’s Distance (EMD) to measure the...

To achieve the full analytical potential of the streaming data from the internet of things, the interconnection of various data sources is needed. By definition, those sources are heterogeneous and their integration is not a trivial task. A common approach to exploit streaming sensor data potential is to use machine learning techniques for predicti...

The availability of open educational resources is growing at an increasingly fast pace since its first promotion by UNESCO in 2002. Today, large variability of opportunities for free and online educational resources are available and accessible by everyone from all around the world who has access to the Internet. An Internet user may exploit number...

Efficient screening of transactions provides an empowering tool for anti-money laundering procedures and actions. Automatic classification and detection of anomalous behaviours and transaction structures enable faster and more effective action on the side of the supervisory authority. This chapter introduces research achievements and tools develope...

Artifical Intelligence (AI) in Education has great potential for building more personalised curricula, as well as democratising education worldwide and creating a Renaissance of new ways of teaching and learning. We believe this is a crucial moment for setting the foundations of AI in education in the beginning of this Fourth Industrial Revolution....

In this paper, we propose a news stream clustering algorithm which directly outputs cross-lingual event clusters. It uses multilingual language models to generate cross-lingual article representations which enable a direct comparison of articles in different languages. The algorithm is evaluated using a cross-lingual news article data set and compa...

Educational recommenders have received much less attention in comparison to e-commerce and entertainment-related recommenders, even though efficient intelligent tutors have great potential to improve learning gains. One of the main challenges in advancing this research direction is the scarcity of large, publicly available datasets. In this work, w...

Educational recommenders have received much less attention in comparison to e-commerce and entertainment-related recommenders, even though efficient intelligent tutors have great potential to improve learning gains. One of the main challenges in advancing this research direction is the scarcity of large, publicly available datasets. In this work, w...

Finding similar documents in a big document corpus based on context has many practical applications especially in the legal sector. In this paper, our focus is on the documents related to environmental law which have been collected in a database of approximately 300k documents. We analyzed the performance of different representation models (called...

In text mining document enrichment processes are used to improve information retrieval. Document enrichment helps us extract metadata from the text which can then be used in document classification. This paper presents the legal domain document enrichment process and analysis of the enriched data. The process of enriching the documents with multipl...

In education we can find different open educational resource (OER) providers that are serving resources in different modalities , formats and languages. These providers can be the actual resource creators or re-distributors that redirect the user to the actual provider. In recent work, we developed a recommendation engine which provides content-bas...

In education we can find millions of video, audio and text educational materials in different formats and languages. This variety and multimodality can impose difficulty on both students and teachers since it is hard to find the right materials that match their learning preferences. This paper presents an approach for retrieving and recommending it...

Todays job market demand from the job seekers to continuously learn new skills. When applying for a job position one must have the required skill set. If the applicant is missing a skill it can be learned by attending a course. Finding the appropriate courses can be tedious but necessary work to be up-to-date with the job market demand. In this pap...

This paper presents learning analytics tools for visual and statistical analysis of data from a portal of video lectures. While learning analytics methods traditionally deal with measurement, collection and analysis of data about learners with an aim of improving the learning process, our solutions are targeted at viewers of VideoLectures.NET. The...