Text, Speech and Dialogue

The International Conference TSD 2003, the sixth event in the series on Text, Speech, and Dialogue, which originated in 1998, presents state-of-the-art technology and - cent achievements in the ?eld of natural language processing. This year s conference includes invited talks given by top-class researchers (Frederick Jelinek from Johns H- kinsUniversityofBaltimore,SadaokiFuruifromtheUniversityofTechnologyinTokyo, President of ISCA, and Steven Krauwer from the Institute of Linguistics of Utrecht U- versity), plenary and problem-oriented sessions, as well as poster sessions and dem- strations involving 28 functional applications. The conference declares its intent to be an interdisciplinary forum, which intertwines researchinspeechandlanguageprocessingaswellasresearchintheEasternandWestern hemispheres. We feel that the mixture of different approaches and applications gives agreatopportunitytogetacquaintedwiththecurrentactivitiesinallaspectsoflanguage communication and to witness the amazing vitality of research from the former East Block countries. The ?nancial support of ISCA (International Speech Communication Association) enables the wide attendance of researchers from all active regions of the world. This volume contains a collection of all the papers presented at the international conf- ence organized by the Faculty of Applied Sciences of the University of West Bohemia in Pilsen in collaboration with the Faculty of Informatics, Masaryk University in Brno, ? and held in the beautiful city of Ceske Budejo ? vice (South Bohemia, Czech Republic), September 8 12, 2003. Each of the submitted papers was thoroughly reviewed by three membersoftheconferencereviewingteamconsistingofmorethan40world-famouss- cialists in the conference topic areas.

Inhalt
Invited Talks.- Combating the Sparse Data Problem of Language Modelling.- Toward Robust Speech Recognition and Understanding.- Text.- A Pilot Study of English Selectional Preferences and Their Cross-Lingual Compatibility with Basque.- Auto-tagging of Text Documents into XML.- Stable Coordinated Pairs in Text Processing.- Backoff DOP: Parameter Estimation by Backoff.- Document Clustering into an Unknown Number of Clusters Using a Genetic Algorithm.- Experiments in German Treebank Parsing.- A Theoretical Basis of an Architecture of a Shell of a Reasonably Robust Syntactic Analyser.- Si3Trenn and Si3Silb: Using the SiSiSi Word Analysis System for Pre-hyphenation and Syllable Counting in German Documents.- Detecting Annotation Errors in a Corpus by Induction of Syntactic Patterns.- The Computational Complexity of Rule-Based Part-of-Speech Tagging.- Text Corpus with Errors.- Identification of Multiwords as Preprocessing for Automatic Extraction of Lexical Similarities.- Build a Large-Scale Syntactically Annotated Chinese Corpus.- Computational Benefits of a Totally Lexicalist Grammar.- Using a Czech Valency Lexicon for Annotation Support.- On Concept Based Approach for Determining Semantic Index Terms.- Corpora Issues in Validation of Serbian Wordnet.- Optimising Attribute Selection in Conversational Search.- Russian Corpus of the 19th Century.- Speech.- On Homogeneous Segments.- Advances in Automatic Speech Recognition by Imitating Spreading Activation.- The Incorporation of Confidence Measures to Language Understanding.- Speech Recognition with ? -Law Companded Features on Reverberated Signals.- Understanding Speech Based on a Bayesian Concept Extraction Method.- A Data-Driven Framework for Intonational Phrase Break Prediction.- Phoneme Recognition Using TemporalPatterns.- Multi-array Multi-speaker Tracking.- Towards Automatic Transcription of Spontaneous Czech Speech in the MALACH Project.- TERSEO: Temporal Expression Resolution System Applied to Event Ordering.- Non-native Pronunciation Variants of City Names as a Problem for Speech Technology Applications.- Improving Speech Recognition by Utilizing Domain Knowledge and Confidence Measures.- Comparison of Acoustic Adaptation Methods in Multilingual Speech Recognition Environment.- A Comparison of Unit Selection Techniques in Limited Domain Speech Synthesis.- Recognition of Speech with Non-random Attributes.- Text-Independent Speaker Verification: The WCL-1 System.- Local Time-Frequency Operators in TRAPs for Speech Recognition.- Entropy and Dynamism Criteria for Speech and Audio Classification Applications.- Speech Production: Phonetic Encoding of Real and Non-words.- Experiments with Automatic Segmentation for Czech Speech Synthesis.- All-Pole Modeling for Definition of Speech Features in Aurora3 DSR Task.- Building of a Vocabulary for the Automatic Voice-Dictation System.- Real-Time Vocal Tract Length Normalization in a Phonological Awareness Teaching System.- Hard-Testing the Multi-stream Approach to Automatic Speech Recognition.- Pitch-Synchronous Speech Signal Segmentation and Its Applications.- Building LVCSR System for Transcription of Spontaneously Pronounced Russian Testimonies in the MALACH Project: Initial Steps and First Results.- The Phase Substitutions in Czech Harmonic Concatenative Speech Synthesis.- Dialogue.- Combining Task Descriptions and Ontological Knowledge for Adaptive Dialogue.- Large Text and Audio Data Alignment for Multimedia Applications.- Building Multilingual Speech Corpora from Interpreted Spontaneous Dialogues on the Net.- Multi-modal Voice Application Design in a Multi-client Environment.- User Modeling and Plan Recognition under Conditions of Uncertainty.- Bilingual Speech Recognition for a Weather Information Retrieval Dialogue System.- On the Use of Prosodic Labelling in Corpus-Based Linguistic Studies of Spontaneous Speech.- A Multi-modal Eliza Using Natural Language Processing and Emotion Recognition.- The Wizard of Oz System for Weather Information Retrieval.- Directives in Estonian Information Dialogues.- SpeechDat-Like Estonian Database.- Dialogue Experiment for Elderly People in Home Health Care System.