Zbigniew Bohdanowicz, Jarosław Kowalski, Katarzyna Abramczuk, Grzegorz Banerski, Daniel Cnotkowski, Agata Kopacz, Paweł Kobyliński, Aldona Zdrodowska, Cezary Biele
2020 W: Human Systems Engineering and Design II : Proceedings of the 2nd International Conference on Human Systems Engineering and Design (IHSED2019): Future Trends and Applications, September 16-18, 2019, Universität der Bundeswehr München, Munich, Germany / Tareq Ahram, Stefan Picki, Redha Taiar, Waldemar Karwowski; Cham: Springer, s. 397-403
The Virtual Reality (VR) world is very suggestive as it intensely affects the senses of vision, hearing and—to a limited extent—touch. It can be expected that in the near future VR will be widely disseminated and used by people of all ages, including children. We decided to conduct a qualitative research project to assess children’s (aged 7–12) first reactions to the use of VR. Children’s opinions and reactions gathered during the interviews indicate that children highly appreciated the attractiveness of the virtual experiences, which were often assessed at a similar or higher level than their real-world counterparts. Our findings clearly suggest that children very easily adopt VR without any prior experience with that technology. We recommend that studies on children’s behavior in VR are continued.
https://link.springer.com/chapter/10.1007/978-3-030-27928-8_61
Marzena Feldy, Marta Bojko
2020 Marketing Instytucji Naukowych i Badawczych, T. 35, z. 1, s. 1-18
Obserwowana od kilku lat na rynku pracy malejąca podaż wykwalifikowanych osób gotowych podjąć zatrudnienie skutkuje umacnianiem się pozycji pracownika. Konsekwencje tego procesu dotykają nie tylko przedsiębiorstwa, ale są również odczuwalne dla instytucji naukowych. Celem artykułu jest diagnoza poziomu zadowolenia z rożnych aspektów pracy i przywiązania do miejsca zatrudnienia w grupach naukowców o odmiennych profilach oczekiwań wobec pracy. Na tej podstawie możliwe będzie wskazanie instytucjom naukowym tych aspektów pracy, o które powinny dbać, aby zapewnić naukowcom wysoki poziom satysfakcji z miejsca zatrudnienia. Aby dostarczyć wiedzy w tym zakresie, analizie poddano materiał empiryczny zgromadzony w ramach badania kwestionariuszowego przeprowadzonego przez OPI PIB w 2017 roku na reprezentatywnej próbie 840 naukowców zatrudnionych we wszystkich typach jednostek naukowych w Polsce. Pracownicy, którzy wzięli udział w sondażu, znajdowali się na różnych etapach kariery naukowej i reprezentowali wszystkie obszary nauki. Na podstawie deklaracji dotyczących oczekiwań wobec pracodawcy przeprowadzono analizę czynnikową i podzielono respondentów na trzy grupy: 1) wymagających, 2) aspirujących i 3) niezaangażowanych. Pracownicy wymagający wyróżniają się wysokimi oczekiwaniami w zakresie wszystkich badanych aspektów pracy: ekonomiczno-organizacyjnych, rozwojowo-społecznych oraz elastyczności zatrudnienia. Z kolei naukowcy aspirujący wyżej niż inne grupy cenią sobie przede wszystkim aspekty rozwojowo-społeczne. Ich przeciwieństwem są zaś pracownicy niezaangażowani, dla których aspekty rozwojowo-społeczne są najmniej ważne, a pozostałe kwestie umiarkowanie istotne. Badanie zadowolenia poszczególnych grup naukowców z obecnego pracodawcy wskazuje na konieczność koncentracji zatrudniających ich instytucji naukowych na odmiennych aspektach pracy. W przypadku pracowników wymagających ważne okazuje się zadbanie o ich dobrostan ekonomiczny. Natomiast dla podniesienia satysfakcji z pracy naukowców z grupy aspirujących istotne będzie zapewnienie im wyższego poziomu zadowolenia ze sfery rozwojowo-społecznej. Największe wyzwanie może stanowić usatysfakcjonowanie pracowników niezaangażowanych, którzy deklarują stosunkowo niski ogólny poziom zadowolenia z miejsca pracy, a jednocześnie nie mają ugruntowanych oczekiwań wobec zatrudniających ich instytucji.
http://minib.pl/oczekiwania-wobec-pracodawcy-a-zadowolenie-z-pracy-naukowcow/
Sławomir Dadas, Michał Perełkiewicz, Rafał Poświata
2020 W: Artificial Intelligence and Soft Computing : 19th International Conference, ICAISC 2020, Zakopane, Poland, October 12-14, 2020, Proceedings, Part II / Leszek Rutkowski, Rafał Scherer, Marcin Korytkowski, Witold Pedrycz, Ryszard Tadeusiewicz, Jacek M. Zurada; Cham: Springer, s. 301-314
19th International Conference, Artificial Intelligence and Soft Computing. Zakopane, 2020-10-12 - 2020-10-14
Transformer-based language models are now widely used in Natural Language Processing (NLP). This statement is especially true for English language, in which many pre-trained models utilizing transformer-based architecture have been published in recent years. This has driven forward the state of the art for a variety of standard NLP tasks such as classification, regression, and sequence labeling, as well as text-to-text tasks, such as machine translation, question answering, or summarization. The situation have been different for low-resource languages, such as Polish, however. Although some transformer-based language models for Polish are available, none of them have come close to the scale, in terms of corpus size and the number of parameters, of the largest English-language models. In this study, we present two language models for Polish based on the popular BERT architecture. The larger model was trained on a dataset consisting of over 1 billion polish sentences, or 135 GB of raw text. We describe our methodology for collecting the data, preparing the corpus, and pre-training the model. We then evaluate our models on thirteen Polish linguistic tasks, and demonstrate improvements over previous approaches in eleven of them.
https://link.springer.com/book/10.1007/978-3-030-61534-5#editorsandaffiliations
Marzena Feldy, Barbara Kowalczyk
2020 European Review, T. 28, z. 4, s. 1-18
The reforms of the Polish system of science emphasise the importance of research and the international activity of Polish scientists. Optimal allocation of funds, taking into account the needs of researchers, is one of the current key objectives of research policy in Poland. The aim of this article is to discover values that Polish researchers identify with as well as defining their approach to the current system of research financing in Poland. The results presented here come from a nationwide study implemented at the turn of 2016. In total, we surveyed 800 randomly selected researchers. The interviewees had at least a PhD degree and were employed in Polish research institutions. The results of the study confirm that the researchers who identify with different ethoses of science have varied approaches to the Polish system of research financing. The scientists who take the least favourable view of the system are those who identify with the academic tradition. Compared with the scientists with deeply-rooted academic values, those in favour of a post-academic ethos more appreciate the available funding opportunities as well as the features of competition-based research funding. They also more often claim that the rules of financing research institutions in Poland have a positive impact on the development of science.
https://www.cambridge.org/core/journals/european-review/article/ethos-of-science-and-the-perception-of-the-polish-system-of-financing-science/BCF6650BC83D0B5515DC9A6E4D6919A9
Sławomir Dadas, Jarosław Protasiewicz, Witold Pedrycz
2019 W: 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC) / Maria Pia Fanti, Mengchu Zhou, Dmitry B. Goldgof, Rodney Roberts; Piscataway, New Jersey: IEEE, s. 3012-3018
2019 IEEE International Conference on Systems, Man and Cybernetics. Bari, 2019-10-06 - 2019-10-09
A model for accurately solving the tasks of intent detection and slot filling requires a substantial amount of user queries annotated manually, which is time-consuming and costly. In this study, in order to circumvent this problem, we propose a method of expanding a training set for handling these two tasks. The data are augmented by applying a random mutation to the training samples, following a set of heuristic rules. For validation of the method, we construct a compound neural architecture composed of a long short-term memory layer, an attention mechanism, and conditional random fields. The experiments conducted on the Automatic Terminal Information Service (ATIS) dataset demonstrate that the models trained on the expanded datasets improve their F-score for the slot-filling task. This improvement is particularly significant for small datasets. We believe that this method allows for the expansion of a small set of previously annotated user queries into a large training set that is sufficient for correctly constructing a model of intent detection and slot filling. Therefore, our approach can be used to save time and money by reducing the amount of data required to train a natural language understanding model. Its novelty consists in expanding a training set for a deep learning model based on a random mutation of training samples, following a set of heuristic rules that utilise external lexicons and the training data itself.
https://ieeexplore.ieee.org/abstract/document/8914542
Łukasz Podlodowski, Marek Kozłowski
2019 W: Proceedings - 2019 IEEE International Conference on Big Data (Big Data) / Roger Barga, Ronay Ak, Kisung Lee, Yuanyuan Tian, Jun Huan, Latifur Khan, Chaitanya Baru, Xiaohua Hu, Yanfang Fanny Ye, Carlo Zaniolo; Piscataway: Institute of Electrical and Electronics Engineers (IEEE), s. 5902-5907
This paper presents an application of XGBoost as a solution for a task associated with the IEEE BigData2019 Cup: Suspicious Network Event Recognition. As has been shown in the paper, the high-quality classification model can be based on independent predictions of each component in the sequence of network traffic events, then analyzed with statistical aggregation functions to generate the final prediction. We also propose the approach to this problem including handling high dimensionality space of IP addresses through encoding octets separately.
https://ieeexplore.ieee.org/document/9006586
Sławomir Dadas
2019 W: Artificial Intelligence and Soft Computing 18th International Conference, ICAISC 2019, Zakopane, Poland, June 16–20, 2019, Proceedings, Part I / Leszek Rutkowski, Rafał Scherer, Marcin Korytkowski, Witold Pedrycz, Ryszard Tadeusiewicz, Jacek M. Zurada; Cham: Springer, s. 39-50
18th International Conference on Artificial Intelligence and Soft Computing. Zakopane, 2019-06-16 - 2019-06-20
Named entity recognition (NER) is one of the tasks in natural language processing that can greatly benefit from the use of external knowledge sources. We propose a named entity recognition framework composed of knowledge-based feature extractors and a deep learning model including contextual word embeddings, long short-term memory (LSTM) layers and conditional random fields (CRF) inference layer. We use an entity linking module to integrate our system with Wikipedia. The combination of effective neural architecture and external resources allows us to obtain state-of-the-art results on recognition of Polish proper names. We evaluate our model on data from PolEval 2018 NER challenge on which it outperforms other methods, reducing the error rate by 22.4% compared to the winning solution. Our work shows that combining neural NER model and entity linking model with a knowledge base is more effective in recognizing named entities than using NER model alone.
https://link.springer.com/book/10.1007/978-3-030-20912-4
Rafał Poświata
2019 W: Proceedings of the 13th International Workshop on Semantic Evaluation / Marianna Apidianaki, Saif M. Mohammad, Jonathan May, Ekaterina Shutowa, Aurelie Herbelot, Zhu Xiaodan; Minneapolis: Association for Computational Linguistics, s. 175-179
13th International Workshop on Semantic Evaluation. Minneapolis, 2019-06-06 - 2019-06-07
This paper describes our system participating in the SemEval-2019 Task 3: EmoContext: Contextual Emotion Detection in Text. The goal was to for a given textual dialogue, i.e. a user utterance along with two turns of context, identify the emotion of user utterance as one of the emotion classes: Happy, Sad, Angry or Others. Our system: ConSSED is a configurable ombination of semantic and sentiment
neural models. The official task submission
achieved a micro-average F1 score of 75.31
which placed us 16th out of 165 participating
systems.
https://www.aclweb.org/anthology/S19-2027/
Michał Perełkiewicz
2019 W: Proceedings of the 13th International Workshop on Semantic Evaluation / Marianna Apidianaki, Saif M. Mohammad, Jonathan May, Ekaterina Shutowa, Aurelie Herbelot, Zhu Xiaodan; Minneapolis: Association for Computational Linguistics, s. 180-184
13th International Workshop on Semantic Evaluation. Minneapolis, 2019-06-06 - 2019-06-07
In this paper, I describe a fusion model combining contextualized and static word representations for approaching the EmoContext task in the SemEval 2019 competition. The model is based on two Recurrent Neural Networks, the first one is fed with a state-of-the-art ELMo deep contextualized word representation and the second one is fed with a static Word2Vec embedding augmented with 10-dimensional affective word feature vector. The proposed model is compared with two baseline models based on a static word representation and a contextualized word representation, separately. My approach achieved officially 0.7278 microaveraged F1 score on the test dataset, ranking 47th out of 165 participants.
https://www.aclweb.org/anthology/S19-2028/
Agata Kopacz, Cezary Biele, Aldona Zdrodowska
2019 Advances in Cognitive Psychology, T. 15, z. 4, s. 256-264
The current study aimed to develop a shortened language-specific (Polish) version of the UNRAVEL task (Altmann, Trafton, & Hambrick, 2014) and to verify whether the adaptation yields valid and reliable data about placekeeping ability. Since the original procedure is intended to investigate task performance referring to placekeeping operations under conditions of task interruptions, we used this tool in the context of a multitasking situation. The adopted version differs from the original
in that we reduced the number of steps in the procedure and changed the rules set, using an acronym WINDA (a word meaning elevator in Polish). Participants were asked to try to keep their place in the WINDA sequence, make a two-alternative forced choice regarding one feature of a presented stimulus, and to continue the task after the interruption at the place where they had left off. Similarly to the original task, reliability of sequence errors was high, suggesting that the WINDA
task is suitable for measuring individual differences in placekeeping performance. The results suggest that the adaptation process that we employed to create the WINDA task can be utilized to generate other language adaptations of this tool (characterized by different levels of difficulty) targeted at specific subject groups.
http://www.ac-psych.org/en/issues/page/9/volume/15/issue/4