Institute of Informatics

Abstract We present experiments and simulations of the electron lithography process aimed at optimizing parameters for the lift-off technique, a critical step in creating metal patterns on a substrate using an electron beam. The study focused on a bilayer consisting of two PMMA resists with different molecular weights - 950,000 on the top and 495,000 on the bottom - corresponding to different sensitivities. The shape of the developed resist images, specifically in the case of a line grating (Line/Space 100/500 nm), was examined. The influence of the exposure dose on the sidewall geometry of the developed profiles in a 550 nm bilayer PMMA resist system was investigated both experimentally and theoretically to achieve nearly vertical profile sidewalls. Contour plots showing the half linewidths of the developed bilayer PMMA resist as a function of the exposure dose along the resist depth were obtained through simulation, allowing for the prediction of profile geometry for doses ranging from 440 μC/cm2 to 760 μC/cm2.

The Webtoon application has become one of the popular platforms for reading comics digitally. Webtoons, as a form of digital comics, present various types of comic content. The success of a Webtoon application depends greatly on understanding the preferences and views of its users. User evaluations of Webtoon applications can provide valuable insight into user satisfaction levels, as well as identify problems that need to be fixed by developers. In this research, Sentiment Analysis was applied to user reviews of the Webtoon Application on the Google Play Store. This research uses two different classification algorithms, namely Naïve Bayes and Random Forest, with the aim of comparing their performance in the context of sentiment analysis of user reviews of Webtoon applications. The results of this research are expected to provide an overview of the most suitable algorithm for conducting sentiment analysis classification in Webtoon applications. In collecting the dataset, we involved webtoon user reviews covering various sentiments, such as positive, negative, and neutral. However, in this analysis, the focus is given to two types of sentiment, namely positive and negative. We apply Naïve Bayes and Random Forest algorithms to perform sentiment classification on the reviews. Performance evaluation is carried out by considering metrics such as accuracy, precision, recall, and F1-score. The results of implementing these two algorithms are an accuracy of 74% Naïve Bayes, and 88% Random Forest. It can be concluded that the Random Forest algorithm is superior to the Naïve Bayes algorithm. With this, the Random Forest algorithm becomes a recommendation for classifying sentiment analysis for Webtoon applications with greater accuracy.

Accurate and precise body measurements are es-sential for various applications, including body composition analysis, health assessment, and clothing design. In comparison to traditional manual methods, 3D scanning has emerged as a faster and more reliable alternative. Our work introduces an automated approach for measuring 3D scanned bodies. Given any 3D mesh representing a human body, we convert it to a template and apply curve-level slicing for the measurements. Our validation incorporates statistical measures proposed for compatibility with the ISO and other methods. Our results improve the previous state-of-the-art in Mean Standard Deviation (MAD) by 5.5% and the Standard Error of Measurements (SEM) by 10.7%.

The amount of antisocial online behavior (AOB) in the political field has been on the rise in recent years. In the research, we are dealing with the reason for confusing AOB with other forms of antisocial behavior. We specifically dealt with NLP in the political field in the Slovak language, with a specific number of prefixes and suffixes. In this paper, we focus on the area of detection of manipulation of political discussions. The data source consisted of previously collected data in the period from April 2018 with an update until November 19, 2022 realized within this research from the site demagog.sk, which we supplemented with new claims. We trained several classification models on preprocessed data and evaluated them. For multi-class classification, the best results were achieved using logistic regression and a support vector machine trained on the resampled dataset—both achieving an accuracy of 0.56 and a macro F1 score of 0.39. In the case of binary classification, the best results were achieved by logistic regression—accuracy 0.7 and macro F1 score 0.56. These models could help detect manipulation in online political discussions.

Theory of Mind constitutes a complex cognitive ability of the individual that impulses him to acquire social knowledge. While the development of Metacognition contributes to the construction of a higher mental structure, allowing the observation, control, and readjustment of the cognitive mechanism with the aim of its optimal performance. The present study investigates the existence of correlation and mutual influence between the two functions. Survey results show that Tom and Metacognition are connected and interact with each other in man's social, cognitive, and evolutionary course, forming two mental processes of outstanding importance for his well-being. Additionally, the predictive power of ToM in the development of Metacognition is highlighted.

Computer Graphics has successfully modeled and rendered many natural patterns, such as the spots on mammals, veins on leaves, and many others. Nevertheless, a fundamental problem remains: how to validate the results beyond subjective visual similarity. We propose a small set of quantitative descriptors tailored to describe the most common pigmentation biological patterns: spots, labyrinths, and stripes, found in many species of mammals and fishes. Our goal is not to classify images blindly, which can be done through deep neural networks but to gain insights into the structure of animal skin pigmentation. For that, we pursue specialized descriptors in the same sense as the general ones from content-based image retrieval systems. First, we compute twelve metrics using image processing tools from a set of real patterns curated from scratch from publicly available repositories. We then normalize those measures by the average and calculate the variance to yield our descriptors. We validate the descriptor’s usefulness through two machine learning tasks on an augmented dataset with real and synthetic patterns. First, the descriptors are used as features of a supervised classifier with an overall accuracy of 98.4%. Second, we test the descriptors in an unsupervised clustering task, differentiating natural from artificial patterns and correctly identifying species present in the natural set. These tests show that the proposed small set of descriptors is representative of the pattern domain analyzed.