AMAgnieszka Mykowiecka

Agnieszka Mykowiecka

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Agnieszka Mykowiecka has a total of 25 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Do transformer-based token classification methods solve the problem of terminology extraction?

    by Małgorzata Marciniak, Piotr Rychlik, Agnieszka Mykowiecka

    Results obtained by transformer-based token classification models are now considered to be a benchmark for the Automatic Terminology Extraction (ATE) task. However, the unsatisfactory results (they rarely exceed 0.7 of the F1 value) raise the question of whether this approach is correct and of what text features are being remembered or inferred by the model trained on this type of annotation. In the paper, we describe a number of experiments using the fine-tuned RoBERTa base model on the ACTER data, RD-TEC, and three Wikipedia articles, which proved that the results of the ATE task obtained by such models depend considerably on the type of texts being processed and their relationship to the training data. While the results are relatively good for some texts with highly specialized vocabulary, the poor results seem to correlate with the high frequency (in general English texts) of tokens that are part of terms in a particular domain. Another property that affects the results is the degree of overlap between the vocabulary of the test data and the vocabulary of terms from the training data. Words that have been labeled as terms in the training data are usually labeled as terms in other, unrelated domains as well. Moreover, we show that the results obtained by these models are unstable—models trained on more data do not include all the items identified by models trained on a smaller dataset and can present substantially lower performance.


  • Neural Nets in Detecting Word Level Metaphors in Polish

    by Aleksander Wawer, Małgorzata Marciniak, Agnieszka Mykowiecka

  • Open Access

    Interpretable segmentation of medical free-text records based on word embeddings

    by Adam Gabriel Dobrakowski, Agnieszka Mykowiecka, Małgorzata Marciniak, Wojciech Jaworski, Przemysław Biecek

    Medical free-text records store a lot of useful information that can be exploited in developing computer-supported medicine. However, extracting the knowledge from the unstructured text is difficult and depends on the language. In the paper, we apply Natural Language Processing methods to process raw medical texts in Polish and propose a new methodology for clustering of patients’ visits. We (1) extract medical terminology from a corpus of free-text clinical records, (2) annotate data with medical concepts, (3) compute vector representations of medical concepts and validate them on the proposed term analogy tasks, (4) compute visit representations as vectors, (5) introduce a new method for clustering of patients’ visits and (6) apply the method to a corpus of 100,000 visits. We use several approaches to visual exploration that facilitate interpretation of segments. With our method, we obtain stable and separated segments of visits which are positively validated against final medical diagnoses. In this paper we show how algorithm for segmentation of medical free-text records may be used to aid medical doctors. In addition to this, we share implementation of described methods with examples as open-source package .


  • Terminology/Keyphrase Extraction for Creation of Book Indexes in Polish

    by Małgorzata Marciniak, Agnieszka Mykowiecka, Piotr Rychlik


  • Side Effect Alerts Generation from EHR in Polish

    by Wojciech Jaworski, Małgorzata Marciniak, Agnieszka Mykowiecka


  • Estimating senses with sets of lexically related words for Polish word sense disambiguation

    by Szymon Rutkowski, Piotr Rychlik, Agnieszka Mykowiecka

    We propose a new algorithm for word sense disambiguation, exploiting data from a WordNet with many types of lexical relations, such as plWordNet for Polish. In this method, sense probabilities in context are approximated with a language model. To estimate the likelihood of a sense appearing amidst the word sequence, the token being disambiguated is substituted with words related lexically to the given sense or words appearing in its WordNet gloss. We test this approach on a set of sense-annotated Polish sentences with a number of neural language models. Our best setup achieves the accuracy score of 55.12% (72.02% when first senses are excluded), up from 51.77% of an existing PageRank-based method. While not exceeding the first (often meaning most frequent) sense baseline in the standard case, this encourages further research on combining WordNet data with neural models.


  • Recognition of irrelevant phrases in automatically extracted lists of domain terms

    by Agnieszka Mykowiecka, Małgorzata Marciniak, Piotr Rychlik

    In our paper, we address the problem of recognition of irrelevant phrases in terminology lists obtained with an automatic term extraction tool. We focus on identification of multi-word phrases that are general terms or discourse expressions. We defined several methods based on comparison of domain corpora and a method based on contexts of phrases identified in a large corpus of general language. The methods were tested on Polish data. We used six domain corpora and one general corpus. Two test sets were prepared to evaluate the methods. The first one consisted of many presumably irrelevant phrases, as we selected phrases which occurred in at least three domain corpora. The second set mainly consisted of domain terms, as it was composed of the top-ranked phrases automatically extracted from the analyzed domain corpora. The results show that the task is quite hard as the inter-annotator agreement is low. Several tested methods achieved similar overall results, although the phrase ordering varied between methods. The most successful method, with a precision of about 0.75 on half of the tested list, was the context based method using a modified contextual diversity coefficient. Although the methods were tested on Polish, they seems to be language independent.


  • Open Access

    Detecting Figurative Word Occurrences Using Recurrent Neural Networks

    by Agnieszka Mykowiecka, Aleksander Wawer, Malgorzata Marciniak


  • Open Access

    Testing word embeddings for Polish

    by Agnieszka Mykowiecka, Małgorzata Marciniak, Piotr Rychlik

    Testing word embeddings for PolishDistributional Semantics postulates the representation of word meaning in the form of numeric vectors which represent words which occur in context in large text data. This paper addresses the problem of constructing such models for the Polish language. The paper compares the effectiveness of models based on lemmas and forms created with Continuous Bag of Words (CBOW) and skip-gram approaches based on different Polish corpora. For the purposes of this comparison, the results of two typical tasks solved with the help of distributional semantics, i.e. synonymy and analogy recognition, are compared. The results show that it is not possible to identify one universal approach to vector creation applicable to various tasks. The most important feature is the quality and size of the data, but different strategy choices can also lead to significantly different results. Testowanie wektorowych reprezentacji dystrybucyjnych słów języka polskiegoSemantyka dystrybucyjna opiera się na założeniu, że znaczenie słów wyrażone jest za pomocą wektorów reprezentujących, w sposób bezpośredni bądź pośredni, konteksty, w jakich słowo to jest używane w dużym zbiorze tekstów. Niniejszy artykuł dotyczy ewaluacji wielu takich modeli skonstruowanych dla języka polskiego. W pracy porównano skuteczność modeli opartych na lematach i formach słów, utworzonych przy wykorzystaniu sieci neuronowych na danych z dwóch różnych korpusów języka polskiego. Ewaluacji dokonano na podstawie wyników dwóch typowych zadań rozwiązywanych za pomocą metod semantyki dystrybucyjnej, tzn. rozpoznania występowania synonimii i analogii między konkretnymi parami słów. Uzyskane wyniki dowodzą, że nie można wskazać jednego uniwersalnego podejścia do tworzenia modeli dystrybucyjnych, gdyż ich skuteczność jest różna w zależności od zastosowania. Najważniejszą cechą wpływającą na jakość modelu jest jakość oraz rozmiar danych, ale wybory różnych strategii uczenia sieci mogą również prowadzić do istotnie odmiennych wyników.