Researchers of the Center for Artificial Intelligence at the FedCSIS conference in Warsaw

On September 17-20, 2023, the FedCSIS conference took place in Warsaw, Poland. The event brought together experts in the field of Artificial Intelligence from different corners of the world. Members of the Center for Artificial Intelligence participated in the event, presenting their innovative work at the conference.

[The presentations will be published in the Annals of Computer Science and Information Systems https://annals-csis.org/volumes (very soon)]

During the AAIA session, a team of researchers from Adam Mickiewicz University Poznan and PWN Scientific Publishers, consisting of Jakub Pokrywka, Krzysztof Jassem, Piotr Wierzchoń (AMU), Piotr Badylak and Grzegorz Kurzyp (PWN), presented an innovative reranking system for a Polish-language medical search engine. This system is designed to support medical professionals by offering them text search based on verified sources.

During the CNLPS + PolEval session, a broad representation from our University, including Marek Kubis, Pawe? Skórzewski, Rafa? Jaworski, Marcin Sowański, Tomasz Zi?tkiewicz and the aforementioned Jakub Pokrywka, Piotr Wierzchoń and Krzysztof Jassem, presented their research on the challenges of conversational AI correctness and punctuation prediction for Polish texts. Meanwhile, Michal Juńczyk presented the concept of BIGOS - a multi-domain benchmark and dataset for Polish Automatic Speech Recognition. Special mention should go to Jakub, who authored or co-authored as many as three presentations. The last one discussed the search and extraction of specific text fragments from large text datasets that are relevant to a given query using the OKAPI BM25 algorithm.

During the same session, Szymon Jadczak, a student in Computer Science and a participant in the AITech project, who won the CAICCAIC challenge organized as part of the conference, was awarded.

In Data Mining Competition, S?awomir Pioronski competed with other participants in a competition to extract valuable information from large data sets, with which he achieved a high third place.

In AI-HuSo session, Piotr Jab?oński presented his paper entitled "Harness Old Media: a cross-disciplinary approach to utilizing television data for media content analysis." This work focused on the use of television data in media content analysis, which is part of a larger project focused on developing algorithms to classify and compare content from different media to identify misinformation.This work is a significant contribution to the development of Artificial Intelligence, especially in the context of Polish language.

Presentations by our representatives demonstrated a deep understanding of the problems and innovative approaches to solving them. The FedCSIS 2023 conference in Warsaw was an excellent platform for sharing knowledge and experience among experts in this rapidly developing field.

Center for Artificial Intelligence website: https://csi.amu.edu.pl

FedCSIC conference website: https://fedcsis.org/