Enabling machines to interact with humans requires understanding what people mean, even when they do not say it explicitly. For example, machines should understand that ``selfish people oppose gun control'' implies a pro-gun control viewpoint (i.e., is taking a stance in support of gun control) despite the negative tone of the statement. Understanding these types of pragmatic inferences allows humans to grasp meaning (e.g., intentions, relevant facts) beyond what is literally expressed in an utterance. Furthermore, pragmatic inferences conveyed through generalizations (e.g., referring to generic ``selfish people'' rather than specific individuals in order to be more persuasive) support flexible and efficient reasoning. Therefore, in this thesis we focus on improving computational understanding of two inter-related types of pragmatic inferences: stancetaking and linguistic generalizations. This thesis is divided into two parts. In Part II, we focus on stance detection. One major challenge for stance detection models is the large and continually growing set of stance targets (i.e., topics to take a stance on). Therefore, to address this we define and study zero-shot stance detection (i.e., evaluation on topics for which there is no training data). Our work develops both datasets and models for this task and analyzes the ongoing challenges for future work. This work has stimulated increasing and ongoing research in zero-shot stance detection in NLP. Then in Part I we study generics --- a specific type of linguistic generalization that does not contain explicit quantifiers (e.g., ``most'', ``some''). These statements can have strong persuasive force and are also related to complex patterns of reasoning. To probe the current understanding capabilities of computational models, we focus on generating generics exemplars --- specific cases when a generic holds true or false. In particular, we propose computational frameworks grounded in linguistic theory to generate the first datasets of exemplars. We then use our datasets to highlight the challenges generics pose for natural language reasoning and the current generic-understanding capabilities of large language models.

The aim of this study is to investigate influence of physical activity on postural balance of children (12 - 15 years old). Method is measurment of postural balance on stabilometric desc in differently challenging positions. Three groups are tested (biathletes, oarsmen and control group), each in count 15 - 20 probands. We expect, young biathletes will have better postural stability than oarsmen and control group thanks to specific training. Keywords Postural balance, static balance, biathlon, rowing, sport, children, laterality, bipedal stance, unipedal stance

No description available.

There is emerging evidence for the potential of occupational back-support exoskeletons (BSEs) to reduce physical demands, and thereby help control/prevent the risk of overexertion injuries associated with manual material handling. However, it is important to understand whether BSEs also introduce any unintended safety challenges. One potential risk associated with BSE use is increased risk of falls, since their extra weight, rigid structure, and external hip extension torque may increase demands on the postural control system. However, there is currently limited evidence on whether, and to what extent, BSE use alters postural stability and/or fall risk. The primary goal of this work was to understand the effects of exoskeleton use, and quantify the effects of exoskeleton design parameters, on balance and postural stability, with a focus on passive BSEs used for repetitive lifting work. A comprehensive evaluation of BSE use was performed under controlled laboratory conditions, focusing on three classes of human activity that form the basis of maintaining postural balance in diverse real-life scenarios: maintenance of a specified posture, voluntary movement, and reaction to an external perturbation. The first study demonstrated that during quiet bipedal stance, BSE use increased median frequency and velocity of the center of pressure in the anterior-posterior direction. In the second study on level walking, BSE use caused an increase in gait step width and gait variability, and decrease in the margin of stability. BSE use with high supportive torque led to adapted gait patterns in early-stance phase. Hip range of motion and peak hip flexion velocity also decreased, and participants exhibited different strategies to increase mechanical energy for propelling the leg in late-stance phase: these effects increased with increasing torque applied by the exoskeleton. In the final study, BSE use did not alter the maximal lean angle from which individuals could successfully execute single step balance recovery, following a forward loss of balance. However, several recovery responses were negatively affected by BSE use, including increased reaction time, impeded hip flexion, and reduced margin of stability in the high-torque condition. This is the first systematical investigation to quantify the effects of passive BSEs with multiple supportive torque levels on balance and postural stability. While exoskeleton effects on static balance were minimal, more substantial changes in gait spatiotemporal parameters, hip joint kinematics, and dynamic margins of stability were observed in the later studies. Our results indicate that postural stability deteriorated with exoskeleton use in dynamic conditions, and provide mechanistic insight into how stability is altered by different exoskeleton design factors such as added mass, restricted range of motion, and external hip extension torque. While our results are suggestive of increased fall risk, especially in the high-torque condition, fall risk in real life is moderated by a complex combination of individual and environmental conditions. Future work should consider more complex, realistic tasks and also include a more diverse sample that is studied under longer exposure durations, to further elucidate these findings. Our characterizations of a wide variety of postural responses as a function of exoskeleton torque settings are expected to contribute to improving both design and practice guidelines to facilitate the safe adoption of BSEs in the workplace. / Doctor of Philosophy / Occupational back-support exoskeletons (BSEs) – wearable mechanical systems designed to support, augment, and/or assist back extension – are expected to serve as an alternative workplace intervention to control and prevent overexertion injuries related to manual material handling tasks. While recent studies have shown the beneficial effects of BSE use in terms of physical load reduction on the low back, some concerns have also been raised on unexpected or unintended effects of exoskeletons. One potential risk associated with exoskeleton use is increased risk of falls, since a BSE's extra weight, rigid structure, and external hip extension torque are expected to place increased demands on the postural control system. Increase in fall risk is a critical safety concern, as occupational falls are a serious problem in terms of injuries, medical/industrial cost, and lost work time. However, there exists limited evidence on whether the use of a BSE alters postural stability and/or increases fall risk. Hence, the goal of our study was to quantify the effects of BSE use on postural stability in various conditions related to real-life scenarios, such as standing balance, walking stability and how one would respond to a loss of balance following an external perturbation. Our results showed that during quiet standing, BSE use slightly increased postural sway. In level walking tasks, BSE use had adverse effects on step length, step width, and dynamic stability. Furthermore, wearing a BSE with high supportive torque led to adapted gait patterns in early-stance phase, whereas participants showed different strategies to increase mechanical energy for propelling the leg in late-stance phase. In the final study investigating single step balance recovery following a forward loss of balance, we found that BSE use negatively affects balance recovery, mainly by impeding hip flexion. Thus, our work suggests that exoskeleton use can deteriorate balance and/or postural stability in situations of static standing, voluntary walking, and reacting to an external perturbation, thereby potentially leading to an increase in fall risk. These effects may be more pronounced among specific population sub-groups such as older workers, and may also affect individuals more severely under conditions of stress or fatigue. Hence, future studies must include more rigorous testing of BSE use using a variety of challenging and realistic scenarios, and also include more diverse population samples. The findings from this work are expected to contribute to improving design and practice guidelines to facilitate the safe adoption of BSEs in the workplace.

The object of this dissertation is hedging in Lithuanian and English research articles in the field of humanities and biomedical sciences. The aim of the research is to identify hedging devices and functions in linguistic and medical research articles in Lithuanian and English and to explore to what extent hedging is language-specific and discipline-specific. As there is a great variety of linguistic resources which convey the pragmatic function of hedging, this research is limited to certain lexical hedges: modal verbs, lexical verbs and adverbials, which are analysed within the conceptual categories of epistemic modality, evidentiality and vagueness. The results of the research suggest that the variety of hedging devices is determined by the discipline, whereas the type of the most frequent lexical hedges depends on the language (English vs. Lithuanian). The frequency of hedges used in the analyzed research articles depends both on the language and discipline. The results of the cross-disciplinary research show that both Lithuanian and English authors of medical texts tend to use fewer hedges than linguists. The variety of hedges employed in the analyzed articles of both disciplines is also different. In both Lithuanian and English linguistic articles a greater variety of hedges has been observed. The results of the cross-linguistic research indicate that according to the main parameters of the analysis, the English authors of the analyzed research articles employ hedges... [to full text] / Disertacijoje tiriamas autoriaus pozicijos švelninimas (angl. hedging) humanitarinių mokslų srities (kalbotyros krypties) ir biomedicinos mokslų srities (medicinos krypties) moksliniuose straipsniuose anglų ir lietuvių kalbomis. Disertacijos tikslas – nustatyti sąšvelnių (angl. hedges) vartosenos tendencijas ir raiškos ypatumus kalbotyros ir medicinos straipsniuose lietuvių ir anglų kalbomis bei išsiaiškinti, kokie veiksniai lemia autoriaus pozicijos raišką tyrinėtuose moksliniuose straipsniuose: konkreti mokslo sritis ar labiau kalbos, kultūros ypatumai. Dėl didelės kalbinių vienetų, galinčių atlikti sąšvelnių funkciją, įvairovės šiame darbe apsiribota tik tam tikromis raiškos priemonėmis: modaliniais veiksmažodžiais, leksiniais veiksmažodžiais ir prieveiksmiais bei aplinkybiniais žodžiais, tirtais iš episteminio modalumo, evidencialumo ir neapibrėžtumo kategorijų perspektyvos. Išanalizavus medžiagą, prieita prie tokių išvadų: sąšvelnių vartosenos įvairovė priklauso nuo disciplinos, o dažniausiai pasirenkamos leksinių vienetų grupės – nuo konkrečios kalbos (anglų vs. lietuvių). Sąšvelnių vartosenos dažnį lemia tiek kalba, tiek disciplina. Tarpdalykinio tyrimo rezultatų palyginimas rodo, kad ir lietuviai, ir anglakalbiai medikai vartoja mažiau tirtų kategorijų sąšvelnių nei kalbininkai. Skiriasi ir vartotų sąšvelnių įvairovė: tirtuose kalbotyros straipsniuose abiejomis kalbomis nustatyta įvairesnių sąšvelnių nei medicinos straipsniuose. Tarpkalbinio tyrimo rezultatai parodė... [toliau žr. visą tekstą]

Title: Methodology of telemark skiing Aim of the study: The aim of this thesis is to create educational, methodological and motivational film, which will include design of metodology of telemark skiing. Methods: In developing this thesis, we used a combination of methods of critical analysis of documents and the author's own empirical experience with telemark skiing with the method of comparison of selected knowledge, methods and participating and uninvolved observation and cinematographic film processing method. The work is a synthesis of theoretical research and dynamic field work. Results: According to this thesis was created teaching methodology and motivational DVD that in addition to the metodology also out lines the historical aspects and current concepts of telemark skiing. The text of the thesis in theoretical part complements the DVD detailed information from the history and information about telemark equipment. Methodology of teaching in the text corresponds to the contents of the DVD. Keywords: telemark, methodology, turn, step, stance, free heel

Title: The comparison of the tennis footwork during the match by the elite female-players. Goals: To describe and analyze tennis footwork from ready position to hit the ball and analyze tennis stance during the match by the elite female-players. Method: In this thesis we collected the data based on observation and videorecording. The data was written sheet of paper down. The obtained data are subjected to final processing using mathematical and mathematical-statistical methods. Results: According to these female-players we have found out that the most footwork to hit the ball is running. During the hit the ball there is no footwork and the most footwork back to ready position is schuffle. The most common tennis stance is open stance. Key words: tennis, tennis footwork, tennis stance

The huge popularity of J.K. Rowling’s Harry Potter novels and Philip Pullman’s challenging trilogy His Dark Materials flagged up a widening audience and increasing status for children’s literature in the West. As Sandra Beckett (2009) notes, children’s fiction is now being embraced with enthusiasm by adult readers, writers, critics and publishers. From this increased profile there has emerged the distinct publishing category of ‘crossover’ fiction. In contrast to earlier children’s novels with broad audience appeal, contemporary crossover works are noted for their contextually radical resistance to conventions and bold innovations in content, style and form. Whilst this has given rise to greater critical interest, however, the focus in general has been on adult authored fiction, rather than the now growing body of work being produced and promoted by children and adolescents themselves. In effect, adult critics and reviewers either exclude or take for granted young authors’ fictions as being formulaic and/or lightweight. The purpose of this study has been to investigate the implications of this stance. Drawing on Mikhail Bakhtin’s (1965) theory of carnival and its associated concepts, I have conducted a comparative analysis of published fiction by adult and teenage authors whose works have been identified as subversive and/or marketed as crossover texts. A Bakhtinian perspective on style, structure and themes in each confirms, or otherwise, their radical status before consideration is given to the implications of any differences in approach. Given John Stephens’ (1999) observation that boundaries between children’s and adults’ fiction are more fundamentally blurred in the fantasy and sub-fantasy modes, the influence of genre has been investigated too. My findings indicate that radical texts with broad audience appeal can, in fact, arise through a variety of narrative forms and writing styles and regardless of authorial age. At the same time, characteristic differences in ‘perspectives’ are shown to mark off adolescent from adult authors’ works. I conclude that the young writers’ near-perspectives can produce hybrid fictions which might be understood as breaking new ground. The fresh insights this study contributes, then, demonstrate that any comprehensive account of the vibrant and ever-shifting contemporary literary scene must encompass broader and altogether more considered critical review of young adults’ input than has been offered to-date.

A dissertação pretende permitir ao leitor simular a forma de pensar de Daniel Dennett, e perpassa toda sua filosofia, com ênfase em seu tratamento de o que são padrões, o algoritmo evolutivo, intuition pumps, consciência, e seu uso dos conceitos de illata, abstracta, semântica e sintaxe para compreender a natureza, a biologia e a mente humana. O trabalho reapresenta, sob nova luz, grande parte das ideias mais importantes de Dennett, e procura fazer a engenharia reversa de o que o levou a pensar de determinadas maneiras, guiando o leitor através de caminhos similares, procurando fomentar um aprendizado ativo de uma forma de pensar, acima e além de uma exposição dos resultados obtidos ao longo de décadas desse pensamento no próprio Dennett / This dissertation intends to provide the reader with an inner simulation of Daniel Dennetts form of reasoning, spreading over his whole philosophy, emphasizing his treatment of patterns, the evolutionary algorithm, consciousness, and his use of illata, abstracta, semantic, and synthax, to carve nature at its joints, especially biology and the human mind. It recasts, in a new light, great part of his most important ideas, and reverse engineers what made him think in particular ways, walking the reader through similar pathways, fostering an active learning of a thinking style, above and beyond a mere exposition of the results obtained by this thinking style over the years

Detecção de posicionamento é a tarefa de automaticamente identificar se o autor de um texto é favorável, contrário, ou nem favorável e nem contrário a uma dada proposição ou alvo. Com o amplo uso do Twitter como plataforma para expressar opiniões e posicionamentos, a análise automatizada deste conteúdo torna-se de grande valia para empresas, organizações e figuras públicas. Em geral, os trabalhos que exploram tal tarefa adotam abordagens supervisionadas ou semi-supervisionadas. O presente trabalho propõe e avalia um processo não supervisionado de detecção de posicionamento em textos de tweets que tem como entrada apenas o alvo e um conjunto de tweets a rotular e é baseado em uma abordagem híbrida composta por 2 etapas: a) rotulação automática de tweets baseada em um conjunto de heurísticas e b) classificação complementar baseada em aprendizado supervisionado de máquina. A proposta tem êxito quando aplicada a figuras públicas, superando o estado-da-arte. Além disso, são avaliadas alternativas no intuito de melhorar seu desempenho quando aplicada a outros domínios, revelando a possibilidade de se empregar estratégias tais como o uso de alvos e perfis semente dependendo das características de cada domínio. / Stance Detection is the task of automatically identifying if the author of a text is in favor of the given target, against the given target, or whether neither inference is likely. With the wide use of Twitter as a platform to express opinions and stances, the automatic analysis of this content becomes of high regard for companies, organizations and public figures. In general, works that explore such task adopt supervised or semi-supervised approaches. The present work proposes and evaluates a non-supervised process to detect stance in texts of tweets that has as entry only the target and a set of tweets to classify and is based on a hybrid approach composed by 2 stages: a) automatic labelling of tweets based on a set of heuristics and b) complementary classification based on supervised machine learning. The proposal succeeds when applied to public figures, overcoming the state-of-the-art. Beyond that, some alternatives are evaluated with the intention of increasing the performance when applied to other domains, revealing the possibility of use of strategies such as using seed targets and profiles depending on each domain characteristics.