Artificial intelligence has achieved enormous progress in recent years and produced highly successful applications in intelligent search, natural language processing and image recognition. Several breakthroughs have been reached in natural language understanding, such as machine translation, speech recognition and generation, semantic search, automatic summarization, etc. To further progress the field, current pure data-driven approaches have to be complemented with additional knowledge sources, such as linguistic, which is the topic of this project. The two teams provide complementary expertise in artificial intelligence (the Slovene partner) and linguistics (the Croatian partner). The results will advance both state-of-the-art in methodology as well as in application of the developed techniques to less-resourced Croatian and Slovene languages. The work will enable highly impactful scientific results and useful, practical contributions to natural language understanding in both Slovene and Croatian, such as word sense disambiguation, relation extraction, and semantic role labeling.