Showing papers on "Czech published in 2009"

Neural network based language models for highly inflective languages

Abstract: Speech recognition of inflectional and morphologically rich languages like Czech is currently quite a challenging task, because simple n-gram techniques are unable to capture important regularities in the data. Several possible solutions were proposed, namely class based models, factored models, decision trees and neural networks. This paper describes improvements obtained in recognition of spoken Czech lectures using language models based on neural networks. Relative reductions in word error rate are more than 15% over baseline obtained with adapted 4-gram backoff language model using modified Kneser-Ney smoothing.

Invisible Religion in a “Non-believing” Country: The Case of the Czech Republic:

Abstract: The authors analyse contemporary ir/religiosity and spirituality in the Czech Republic using data from national censuses, international surveys and a specialized national survey on the de-traditionalization and individualization of religion from 2006. The authors conclude that the influence of socio-demographic characteristics on a respondent’s religiosity, both traditional and alternative, is weak; a more important factor in the determination of a person’s world-view was found to be his/her religious socialization or the absence thereof. The effectiveness of religious socialization differs significantly between the various churches; the least effective being the Catholic Church and Czechoslovak Hussite Church. For historical reasons, attitudes towards Catholicism also influence the Czech mind-set on religion in general. The out-of-church movement became established very early in the country’s history and was subsequently strengthened by Communist anticlericalism resulting in today’s prevalence of “non-be...

Czech Named Entity Corpus and SVM-based Recognizer

Abstract: This paper deals with recognition of named entities in Czech texts. We present a recently released corpus of Czech sentences with manually annotated named entities, in which a rich two-level classification scheme was used. There are around 6000 sentences in the corpus with roughly 33000 marked named entity instances. We use the data for training and evaluating a named entity recognizer based on Support Vector Machine classification technique. The presented recognizer outperforms the results previously reported for NE recognition in Czech.

The International Corpus of Learner English. Version 2. Handbook and CD-Rom

Abstract: The International Corpus of Learner English (Version 2) is a corpus of writing by higher intermediate to advanced learners of English. It contains 3.7 million words of EFL writing from learners representing 16 different mother tongue backgrounds (Bulgarian, Chinese, Czech, Dutch, Finnish, French, German, Italian, Japanese, Norwegian, Polish, Russian, Spanish, Swedish, Turkish and Tswana). It differs from the first version published in 2002 not only by its increased size and range of learner populations, but also by its interface, which contains two new functionalities: built-in concordancer allowing users to search for word forms, lemmas and/or part-of-speech tags and breakdown of the query results according to the learner profile information.

Neo‐liberalism by Decay? The Evolution of the Czech Welfare State

Abstract: In this article we analyse the evolution of the Czech welfare state and we examine the factors explaining its path. We show that although the Czech welfare regime exhibits a 'mixed profile' that in ...

Syntactic analysis using finite patterns: a new parsing system for Czech

Abstract: Syntactic analysis of natural languages is considered to be one of the basic steps to advanced natural language processing, such as logical analysis or information retrieval with natural language texts. The Czech language can be characterized as a morphologically rich language with a relatively free word order, which further complicates the problem of syntactic analysis. Current parsing systems for Czech fight many problems including low precision or high ambiguity of the parser output. In this paper, we show a new approach to syntactic analysis of free-word-order languages based on the idea of pattern matching linking rules. The system, named SET, is currently developed and tested with the Czech language as a representative of free-word-order languages with very rich morphological system. We briefly mention current approaches and parsing systems for Czech. Then we describe the basic ideas as well as details of SET's prototype implementation of the pattern matching approach to syntactic analysis. We also offer preliminary analysis of the system parsing precision and discuss the advantages and disadvantages of this approach.

Cross-Modal Analysis of Speech, Gestures, Gaze and Facial Expressions: COST Action 2102 International Conference Prague, Czech Republic, October 15-18, 2008 Revised Selected and Invited Papers

Abstract: Emotions and ICT.- Cross-Fertilization between Studies on ICT Practices of Use and Cross-Modal Analysis of Verbal and Nonverbal Communication.- Theories without Heart.- Prosodic Characteristics and Emotional Meanings of Slovak Hot-Spot Words.- Affiliations, Emotion and the Mobile Phone.- Polish Emotional Speech Database - Recording and Preliminary Validation.- Towards a Framework of Critical Multimodal Analysis: Emotion in a Film Trailer.- Biosignal Based Emotion Analysis of Human-Agent Interactions.- Emotional Aspects in User Experience with Interactive Digital Television: A Case Study on Dyslexia Rehabilitation.- Investigation of Normalised Time of Increasing Vocal Fold Contact as a Discriminator of Emotional Voice Type.- Evaluation of Speech Emotion Classification Based on GMM and Data Fusion.- Spectral Flatness Analysis for Emotional Speech Synthesis and Transformation.- Verbal and Nonverbal Features of Computational Phonetics.- Voice Pleasantness of Female Voices and the Assessment of Physical Characteristics.- Technical and Phonetic Aspects of Speech Quality Assessment: The Case of Prosody Synthesis.- Syntactic Doubling: Some Data on Tuscan Italian.- Perception of Czech in Noise: Stability of Vowels.- Challenges in Segmenting the Czech Lateral Liquid.- Implications of Acoustic Variation for the Segmentation of the Czech Trill /r/.- Voicing in Labial Plosives in Czech.- Normalization of the Vocalic Space.- Algorithmic and Theoretical Analysis of Multimodal Interfaces.- Gaze Behaviors for Virtual Crowd Characters.- Gestural Abstraction and Restatement: From Iconicity to Metaphor.- Preliminary Prosodic and Gestural Characteristics of Instructing Acts in Polish Task-Oriented Dialogues.- Polish Children's Gesticulation in Narrating (Re-telling) a Cartoon.- Prediction of Learning Abilities Based on a Cross-Modal Evaluation of Non-verbal Mental Attributes Using Video-Game-Like Interfaces.- Automatic Sentence Modality Recognition in Children's Speech, and Its Usage Potential in the Speech Therapy.- Supporting Engagement and Floor Control in Hybrid Meetings.- Behavioral Consistency Extraction for Face Verification.- Protecting Face Biometric DCT Templates by Means of Pseudo-random Permutations.- Facial Expressions Recognition from Image Sequences.- Czech Artificial Computerized Talking Head George.- An Investigation into Audiovisual Speech Correlation in Reverberant Noisy Environments.- Articulatory Speech Re-synthesis: Profiting from Natural Acoustic Speech Data.- A Blind Source Separation Based Approach for Speech Enhancement in Noisy and Reverberant Environment.- Quantitative Analysis of the Relative Local Speech Rate.- Czech Spontaneous Speech Collection and Annotation: The Database of Technical Lectures.- BSSGUI - A Package for Interactive Control of Blind Source Separation Algorithms in MATLAB.- Accuracy Analysis of Generalized Pronunciation Variant Selection in ASR Systems.- Analysis of the Possibilities to Adapt the Foreign Language Speech Recognition Engines for the Lithuanian Spoken Commands Recognition.- MLLR Transforms Based Speaker Recognition in Broadcast Streams.

Religion: An Unsolved Problem for the Modern Czech Nation

Abstract: The Czech Republic is widely known as ‘the least religious’ country in the world and most Czechs are quite proud of that fact. The authors, however, challenge both of these characteristics. Czechs might better be considered unchurched than atheist, with various forms of modern New Age spirituality steadily gaining in popularity. Moreover, their reputation for irreligiosity is somewhat questionable, since it is most often based upon communist (and other more historically deep-rooted) anticlerical notions, while people have little real knowledge of the ideas which they so readily reject. These assertions are based both on quantitative data, provided by census returns and ISSP surveys on religion, and on qualitative data, collected in local ethnographic research in the town of Česká Lípa in northern Bohemia, designed along the lines of the Lancaster University Kendal Project in Great Britain. The Czech population can be divided into three ‘blocks’, religionists, spiritualists, and atheists/unbelievers, none of which, however, can be considered uniform in terms of membership or truly mutually exclusive. The authors conclude that traditional religionists of various denominations, the followers of New Age movements, and the ‘rest’ of the population can be seen as three distinctive groups within society and that mutual understanding and acceptance are by no means the norm.

Czech, German and English: Finding Their Place in Multinational Companies in the Czech Republic

Abstract: Large multinational companies/corporations (multinationals) operate in several countries and are characteristic of the exchange of know-how, people and products which takes place between parent and daughter companies (subsidiaries). An important common feature of such businesses is that several operational languages are involved. This chapter examines the changing roles of three languages, Czech, German and English, in the everyday functioning of Czech-located subsidiaries/branches of companies based in Germany. It identifies some of the past and present language problems in the multinationals and how these problems are managed, particularly in relation to the local conditions and practices and the socio-economic structure, with respect to the relationship between the old and new EU member states. The chapter is based on long-term research of such companies operating in the Czech Republic (see Nekula et al. 2005;Nekvapil and Nekula 2006a, 2006b), but it extends this research, drawing on more recent fieldwork and new conceptualizations.1 We focus on the situation in one multinational, but view this multinational against the background of other companies.

A history of the Czech Lands

Abstract: Born January 1, 1993, after it split with Slovakia, the Czech Republic is one of the youngest members of the European Union. Despite its youth as a nation, this land and the areas just outside its modern borders boasts an ancient and intricate past. With "A History of the Czech Lands", editors Jaroslav Panek and Oldrich Tuma - along with several scholars from the Academy of Sciences of the Czech Republic and Charles University Prague - provide one of the most complete historical accounts of this region to date. Panek and Tuma's history begins in the Neolithic era and follows the development of the state as it transformed into the Kingdom of Bohemia during the ninth century, into Czechoslovakia after World War I, and finally into the Czech Republic. Such a tumultuous political past arises in part from a fascinating native people, and "A History of the Czech Lands" profiles the Czechs in great detail, delving into past and present traditions and explaining how generation after generation adapted to a perpetually changing government and economy. In addition, the contributors examine the many minorities that now call these lands home - Jews, Slovaks, Poles, Germans, Ukrainians, and others - and how each group's migration to the region has contributed to life in the Czech Republic today. The first study in English with this scope and ambition, "A History of the Czech Lands" is essential for scholars of Slavic, Central, and East European studies and a must-read for those who trace their ancestry to these lands.

Challenges in speech processing of slavic languages (case studies in speech recognition of czech and slovak)

Abstract: Slavic languages pose a big challenge for researchers dealing with speech technology. They exhibit a large degree of inflection, namely declension of nouns, pronouns and adjectives, and conjugation of verbs. This has a large impact on the size of lexical inventories in these languages, and significantly complicates the design of text-to-speech and, in particular, speech-to-text systems. In the paper, we demonstrate some of the typical features of the Slavic languages and show how they can be handled in the development of practical speech processing systems. We present our solutions we applied in the design of voice dictation and broadcast speech transcription systems developed for Czech. Furthermore, we demonstrate how these systems can be converted to another similar Slavic language, in our case Slovak. All the presented systems operate in real time with very large vocabularies (350K words in Czech, 170K words in Slovak) and some of them have been already deployed in practice.

For Business, for Pleasure or for Necessity? The Czech Republic's Choices for Europe

Abstract: Drawing on scholarly accounts from the study of Western Europe which seek to explain national preference formation in European Union member states, this article explores the factors which shape the Czech Republic's policy preferences at the EU level. It argues that whilst ideology, dependency and powerful economic interest groups have explanatory power, the key to explaining national preference formation in the Czech case lies in vulnerabilities and (perceived) weakness.

Evaluation of Machine Translation Metrics for Czech as the Target Language

Abstract: In the present work we study semi-automatic evaluation techniques of machine translation (MT) systems. These techniques are based on a comparison of the MT system’s output to human translations of the same text. Various metrics were proposed in the recent years, ranging from metrics using only a unigram comparison to metrics that try to take advantage of additional syntactic or semantic information. The main goal of this article is to compare these metrics with respect to their correlation with human judgments for Czech as the target language and to propose the best ones that can be used for an evaluation of MT systems translating into Czech language.

Comparative assessment of young learners’ foreign language competence in three Eastern European countries

Abstract: This paper concerns teacher practices in, and beliefs about, the assessment of young learners’ progress in English in three Eastern European countries (Slovenia, Croatia, and the Czech Republic). The central part of the paper focuses on an international project involving empirical research into assessment of young learners’ foreign language competence in Slovenia, Croatia and the Czech Republic. With the help of an adapted questionnaire, we collected data from a non‐random sample of primary and foreign language teachers who teach foreign languages at the primary level in these countries. The research shows that English as a foreign language is taught mostly by young teachers either primary specialists or foreign language teachers. These teachers most frequently use oral assessment/interviews or self‐developed tests. Other more authentic types of assessment, such as language portfolios, are rarely used. The teachers most frequently assess speaking and listening skills, and they use assessment involving voc...

Research of the Use of Agile Methodologies in the Czech Republic

Abstract: Agile methodologies have recently been widely gaining ground. We assumed another situation in the Czech Republic. Therefore we decided to conduct our own research based on a survey. The research objective was to determine the rate of agile approaches usage and practical experience with these approaches in companies in the Czech Republic. This paper presents the results of that research.

Choosing Slovakia: Slavic Hungary, the Czechoslovak Language and Accidental Nationalism

Abstract: At the turn of the nineteenth century, Hungary was the site of a national awakening. While Hungarian-speaking Hungarians sought to assimilate Hungary's ethnic minorities into a new idea of nationhood, the country's Slavs instead imagined a proud multi-ethnic and multi-lingual state whose citizens could freely use their native languages. The Slavs saw themselves as Hungarian citizens speaking Pan-Slav and Czech dialects - and yet were the origins of what would become in the twentieth century a new Slovak nation. How then did Slovak nationalism emerge from multi-ethnic Hungarian loyalism, Czechoslovakism and Pan-Slavism? Here Alexander Maxwell presents the story of how and why Slovakia came to be.

The academic voice in English and Czech higher education quality

Abstract: Purpose – This paper sets out to report on a research project investigating the academic voice in higher education quality in the UK and the Czech Republic. It aims to describe the origins and reasons for introducing quality monitoring and assurance into higher education, showing the differences and impacts on higher education quality in England and the Czech Republic, including the current practices and presenting the concerns and issues voiced by the academics and higher education leaders in both higher education systems.Design/methodology/approach – The research utilised a critical event narrative inquiry method, which focuses on issues of complexity and human‐centredness in studied phenomena. In this way the method addresses issues that are frequently overlooked by quantitative research methods. It is argued that, by extracting “critical events,” the method is more efficient in dealing with large amounts of data, which often result from the use of qualitative research methods. In the presented researc...

Rodiče na trhu práce: mezi prací a péčí

Abstract: In this article the authors map the theoretical arguments on the gender dimension of the welfare state. They propose three integral dimensions of conceptualising the exercise of parenthood in Czech society in relation to gender equality in the labour market that co-determine the position of parents in the labour market. The authors analyse these dimensions using data from two representative sociological surveys. 1) The right to be a parent (to care for one's child) and the right to work: the measures provided in the Czech welfare state are based on the myth that there are two separate worlds of work and care in conformity with the gender principle, even though there are no significant differences between Czech men and women in terms of the value of work in their lives. 2) Equality or non-discrimination in parenthood: the right to work of mothers of young children is violated in the context of generally increasing gender inequalities in the labour market. 3) The opportunity to achieve a work/life balance: in Czech society flexible forms of employment are uncommon, working hours tend to have a fixed start and finish, or there is negative flexibility, which renders a work/life balance impossible. The way in which state policy defines and employers apply the conditions of parenthood in relation to the labour market and in the context of the gender structure of Czech society makes parenthood a significant handicap for the social inclusion of women who are mothers of young children in the Czech Republic.

Using Morphological Information for Robust Language Modeling in Czech ASR System

Abstract: Automatic speech recognition, or more precisely language modeling, of the Czech language has to face challenges that are not present in the language modeling of English. Those include mainly the rapid vocabulary growth and closely connected unreliable estimates of the language model parameters. These phenomena are caused mostly by the highly inflectional nature of the Czech language. On the other hand, the rich morphology together with the well-developed automatic systems for morphological tagging can be exploited to reinforce the language model probability estimates. This paper shows that using rich morphological tags within the concept of class-based n-gram language model with many-to-many word-to-class mapping and combination of this model with the standard word-based n-gram can improve the recognition accuracy over the word-based baseline on the task of automatic transcription of unconstrained spontaneous Czech interviews.

Time to death and health expenditure of the Czech health care system

Abstract: Growing concern about future sustainability of public budgets in the context of population ageing has given rise to a large debate on the role of age in the context of health care expenditure. Growing evidence on the so called death related costs hypothesis arguing that the positive relationship between age of the cohort and related health care expenditure is the result of growing probability of death changes in an important manner the results of the projections. The aim of this paper is to explore the importance of the death related costs hypothesis in the Czech health expenditure data and the impact of the hypothesis on the projection of the financial sustainability of the Czech health care system.

Toward the Economy of Pedestrian Movement in Czech and Slovak Shopping Malls

Abstract: This article examines changes in shopping behavior in two postcommunist countries, the Czech Republic and Slovakia, by replicating Bitgood and Dukes's study of pedestrian choice point behavior. It ...

Comparative study on circadian typology and sleep habits of Japanese and Czech infants aged 0–8 years

Abstract: Average bedtime and sleep duration of Japanese infants aged 0–8 years were later by 1.3 h and shorter by 1 h, respectively, than those of Czech infants. Japanese infants were predominantly more evening-type than Czech infants, while Japanese parents (mostly mothers) were significantly more morning-type than Czech mothers. Correlation value (r-value) between the morningness-eveningness (M-E) scores of infants and mothers was 0.405 in Japanese participants, whereas it was relatively low (0.297) in Czech ones. Czech infants were reported as being depressed more frequently than Japanese ones who were reported as being more frequently angry than the Czech infants.

English-Czech MT in 2008

Abstract: We describe two systems for English-to-Czech machine translation that took part in the WMT09 translation task. One of the systems is a tuned phrase-based system and the other one is based on a linguistically motivated analysis-transfer-synthesis approach.

The politics of language and nationalisms in modern Central Europe

Abstract: Foreword P.Burke Author Preface Introduction Language in Central Europe: An Overview The Broader Linguistic and Cultural Context of Central Europe PART I: CENTRAL EUROPEAN POLITICS AND LANGUAGES IN THE LONG NINETEENTH CENTURY The Polish Case: From Natio to Nation The Hungarian Case: From Natio to the Ersatz Nation-state The Czech Case: From the Bohemian Slavophone Populus to Czech Nationalism and the Czechoslovak Nation The Slovak Case: From Upper Hungary's Slavophone populus to Slovak nationalism and the Czechoslovak nation PART II: NATIONALISMS AND LANGUAGE IN THE SHORT TWENTIETH CENTURY The Polish Nation: From a Multiethnic to an Ethnically Homogenous Nation-State The Hungarian Nation: From Hungary to Magyarorszag The Czech Nation: Between Czechoslovak and Czech Nationalism The Slovak Nation: From Czechoslovakia to Slovakia, Conclusion Bibliography

Czech and Slovak Cinema: Theme and Tradition

Abstract: 1 Introduction 2 The Comic Tradition 3 Czech Lyricism 4 The Avant Garde 5 Surrealism 6 Animation 7 Representations of the Holocaust 8 Realism and the New Wave 9 Politics and Film 10 Traditions of the Absurd 11 Is There a Slovak Style?