PolEval 2022/23 papers and presentations

PolEval 2022/23 results were presented at FedCSIS 2023, in a special session of the 1st Symposium on Challenges for Natural Language Processing (CNLPS’23) on September 20, 2023 in Warsaw.

PolEval 2022/23 overview

• Tasks and Results (Łukasz Kobyliński, Maciej Ogrodniczuk)

Task 1: Punctuation prediction from conversational language

• Task overview (Piotr Pęzik, Adam Wawrzyński, Michał Adamczyk, Sylwia Karasińska, Wojciech Janowski, Agnieszka Mikołajczyk, Filip Żarnecki, Patryk Neubauer, Anna Cichosz)
• How to predict punctuation: Quick guide (Oskar Bujacz)
• Punctuation Prediction for Polish Texts using Transformers (Jakub Pokrywka)

Task 2: Abbreviation disambiguation

• Task overview (Michał Marcińczuk, Łukasz Kobyliński)
• Abbreviation Disambiguation in Polish Press News Using Encoder-Decoder Models (Krzysztof Wróbel, Jakub Karbowski, Paweł Lewkowicz)

Task 3: Passage retrieval

• Task overview (Piotr Rybak, Aleksander Smywiński-Pohl, Piotr Przybyła, Maciej Ogrodniczuk)
• Passage Retrieval of Polish Texts Using OKAPI BM25 and an Ensemble of Cross Encoders (Jakub Pokrywka)
• Hybrid retrievers with generative re-rankers (Marek Kozłowski)
• Multi-index Retrieve and Rerank with Sequence-to-Sequence Model (Konrad Wojtasik)
• Passage Retrieval in question answering systems in Polish language (Anna Pacanowska)

The papers were published in the Proceedings of the 2023 Federated Conference on Computer Science and Intelligence Systems:

• PolEval 2022/23 Challenge Tasks and Results (Łukasz Kobyliński, Maciej Ogrodniczuk, Piotr Rybak, Piotr Przybyła, Piotr Pęzik, Agnieszka Mikołajczyk, Wojciech Janowski, Michał Marcińczuk, Aleksander Smywiński-Pohl)
• Punctuation Prediction for Polish Texts using Transformers (Jakub Pokrywka)
• Abbreviation Disambiguation in Polish Press News Using Encoder-Decoder Models (Krzysztof Wróbel, Jakub Karbowski, Paweł Lewkowicz)
• Passage Retrieval of Polish Texts Using OKAPI BM25 and an Ensemble of Cross Encoders (Jakub Pokrywka)
• Hybrid retrievers with generative re-rankers (Marek Kozłowski)
• Multi-index Retrieve and Rerank with Sequence-to-Sequence Model (Konrad Wojtasik)
• Passage Retrieval in question answering systems in Polish language (Anna Pacanowska)