Text, Speech, and Dialogue : 18th International Conference, TSD 2015, Pilsen,Czech Republic, September 14-17, 2015, Proceedings / / edited by Pavel Král, Václav Matoušek

• Pubblicazione: Cham : , : Springer International Publishing : , : Imprint : Springer, , 2015
• Edizione: 1st ed. 2015.
• Descrizione fisica: 1 online resource (XVIII, 612 p. 122 illus. in color.)
• Disciplina: 006.35
• Soggetto topico: Artificial intelligence; Application software; Information storage and retrieval; Database management; Pattern recognition; Data mining; Artificial Intelligence; Information Systems Applications (incl. Internet); Information Storage and Retrieval; Database Management; Pattern Recognition; Data Mining and Knowledge Discovery
• Persona (resp. second.): KrálPavel; MatoušekVáclav
• Note generali: Includes index.
• Nota di contenuto: Intro -- Preface -- Organization -- About Plzeň (Pilsen) -- Contents -- Invited Talks -- Speech Analysis in the Big Data Era -- 1 Introduction -- 2 Data: The Availability-Shock -- 3 On Efficiency: Learning Cooperatively -- 3.1 Transfer Learning -- 3.2 (Dynamic) Active Learning -- 3.3 Semi-supervised Learning -- 3.4 Cooperative Learning -- 4 On Decision-Making: Learning Confidence Measures -- 4.1 Agreement-Based Confidence Measures -- 4.2 Learning Errors -- 5 On Seeing the Larger Picture: Learning Multiple Targets -- 6 On Big Data: Distribution -- 7 Conclusion -- References -- I Conference Papers -- A Multi-criteria Text Selection Approach for Building a Speech Corpus -- 1 Introduction -- 2 Literature Review -- 3 Proposed Approach -- 4 Experimental Results -- 5 Conclusion -- References -- Experiment with GMM-Based Artefact Localization in Czech Synthetic Speech -- 1 Introduction -- 2 Method -- 3 Material, Experiments, and Results -- 4 Discussion and Conclusion -- References -- Tuned and GPU-Accelerated Parallel Data Mining from Comparable Corpora -- 1 Introduction -- 2 State of the Art -- 3 Parallel Data Mining -- 4 Yalign and Improvements -- 5 Evaluation of Obtained Comparable Corpora -- 6 Conclusions -- References -- Investigating Genre and Method Variation in Translation Using Text Classification -- 1 Introduction -- 2 Related Work and Theoretical Background -- 3 Methods -- 3.1 Data -- 3.2 Algorithms -- 4 Results -- 4.1 Genres and Methods -- 4.2 Translation Methods -- 4.3 Different Genres: Different Language? -- 4.4 Human vs. Machine -- 4.5 Feature Analysis -- 5 Conclusion and Outlook -- References -- Extracting Characteristics of Fashion Models from Magazines for Item Recommendation -- 1 Introduction -- 2 Previous Research -- 3 Proposed Method -- 3.1 Acquiring Item Name and Description -- 3.2 Finding the Features of the Model from an Item Name. ; 3.3 Finding the Feature of the Model from the Item Description -- 3.4 Creating a Fashion Style Vector -- 4 Experiment -- 4.1 Method 1 Data -- 4.2 Method 2 Data -- 4.3 Image Score -- 4.4 Evaluation -- 4.5 Evaluation Results -- 4.6 Discussion -- 5 Conclusion -- References -- Segment Representations in Named Entity Recognition -- 1 Introduction -- 2 Segment Representations -- 3 Related Work -- 4 NER System -- 5 Corpora -- 6 Experiments -- 6.1 Standard Partitioning -- 6.2 Significance Tests -- 6.3 Discussion -- 7 Conclusion -- References -- Analyzing Text Coherence via Multiple Annotation in the Prague Dependency Treebank -- 1 Introduction -- 2 Aim of Work -- 3 Language Material -- the Prague Dependency Treebank -- 3.1 Annotation of Sentence Information Structure -- 3.2 Annotation of Bridging Anaphora -- 4 Methods -- 5 Results -- 6 Conclusion -- References -- Automatic Detection of Parkinson's Disease in Reverberant Environments -- 1 Introduction -- 2 Experimental Setup -- 2.1 Databases -- 2.2 Speech Tasks -- 2.3 Reverberation -- 2.4 Preprocessing and Characterization of Unvoiced Frames -- 2.5 Classification -- 3 Results and Discussion -- 4 Conclusions -- References -- Automatic Detection of Parkinson's Disease from Compressed Speech Recordings -- 1 Introduction -- 2 Experimental Setup -- 2.1 Speech Recordings -- 2.2 Encoding - Compression -- 2.3 Pre-processing and Voiced/Unvoiced Segmentation -- 2.4 Characterization -- 2.5 Classification -- 3 Results -- 4 Conclusion -- References -- Time Dependent ARMA for Automatic Recognition of Fear-Type Emotions in Speech -- 1 Introduction -- 2 Materials and Methods -- 2.1 Segmentation -- 2.2 SP-TARMA Modeling -- 2.3 Feature Estimation -- 2.4 Classification -- 3 Experimental Framework and Results -- 3.1 Datasets -- 3.2 Experimental Setup -- 3.3 Results and Discussion -- 4 Conclusion -- References. ; Using Lexical Stress in Authorship Attribution of Historical Texts -- 1 Introduction -- 2 Lexical Stress -- 2.1 Motivation -- 2.2 Extracting Lexical Stress from Text -- 3 Machine Learning Algorithms -- 3.1 Learning Methods -- 3.2 Features Used -- 3.3 Evaluating Performance -- 3.4 Weighted Voting -- 4 Authorship Attribution with Lexical Stress Pattern Vectors -- 4.1 Experiments with Lexical Stress Patterns Only -- 4.2 Combining Stress with Other Lexical Features -- 5 Conclusion and Future Work -- References -- Eye Gaze Analyses in L1 and L2 Conversations: Difference in Interaction Structures -- 1 Introduction -- 2 Mutimodal Corpus -- 2.1 Participants -- 2.2 Experimental Setup -- 2.3 Procedure -- 2.4 Annotations -- 2.5 Transcription -- 3 Analyses -- 3.1 Analysis 1: Listener's Eye Gaze Activities -- 3.2 Analysis 2: Speaker's Eye Gaze Activities -- 3.3 Analysis 3: Eye Gaze Activities Among the Participants -- 4 Discussion -- 5 Conclusion -- References -- Word Categorization of Corporate Annual Reports for Bankruptcy Prediction by Machine Learning Methods -- 1 Introduction -- 2 Data Preprocessing -- 3 Experimental Results -- 4 Conclusion -- References -- Novel Multi-word Lists for Investors' Decision Making -- 1 Introduction -- 2 Research Methodology -- 3 Data Collection and Description -- 4 Relationship Between Word Lists and Investment Indicators -- 5 Conclusion -- References -- Topic Classifier for Customer Service Dialog Systems -- 1 Introduction -- 2 The Task and Corpus -- 3 Semantic Parsing as a Previous Step to Classification -- 3.1 The Semantic Grammar -- 3.2 Word-Level Grammars -- 4 Entropy Modeling -- 5 Classification -- 6 Experiments and Results -- 7 Conclusions and Future Work -- References -- Defining a Global Adaptive Duration Target Cost for Unit Selection Speech Synthesis -- 1 Introduction -- 2 The TTS System. ; 3 An Adaptive Duration Target Cost -- 3.1 Neural Network -- 3.2 Duration Target Cost -- 4 Experiments -- 4.1 Corpus Description -- 4.2 Objective Analysis -- 4.3 Subjective Evaluation -- 5 Conclusion -- References -- Adaptive Speech Synthesis of Albanian Dialects -- 1 Introduction -- 2 The Albanian Language -- 3 Grapheme-to-Phoneme Conversion -- 4 Recording Script -- 5 Recording -- 6 Voice Building -- 7 Evaluation -- 7.1 Design -- 7.2 Results -- 8 Conclusion -- References -- Language-Independent Age Estimation from Speech Using Phonological and Phonemic Features -- 1 Introduction -- 2 Test Data and Subjective Evaluation -- 3 Features Computed from the Speech Data -- 4 Results and Discussion -- References -- LecTrack: Incremental Dialog State Tracking with Long Short-Term Memory Networks -- 1 Introduction -- 2 Dialog State Tracking -- 3 LSTM Dialog State Tracker -- 3.1 Model -- 3.2 Training -- 4 Experiments -- 4.1 Dataset -- 4.2 Baseline -- 4.3 Data Preprocessing -- 4.4 Experimental Methodology -- 4.5 Results -- 5 Discussion -- 6 Related Work -- 7 Conclusion -- References -- Automatic Construction of Domain Specific Sentiment Lexicons for Hungarian -- 1 Introduction -- 2 Sentiment Lexica -- 2.1 Translating a Foreign Lexicon -- 2.2 Bootstrapping Sentiment Lexicon -- 2.3 Extending Lexicons -- 3 Data -- 4 Results -- 5 Conclusions -- References -- Investigation of Word Senses over Time Using Linguistic Corpora -- 1 Introduction -- 2 Related Work -- 3 Topic Models -- 4 Topic Models over Time -- 5 Experiments -- 6 Conclusion and Future Work -- References -- Dependency-Based Problem Phrase Extraction from User Reviews of Products -- 1 Introduction -- 2 Related Work -- 3 Target Phrase Extraction -- 3.1 Dependency Relations for Target Extraction -- 3.2 Calculating Semantic Relatedness of Problem Targets -- 3.3 Dependency-Based Approach. ; 4 Evaluation and Experiments -- 4.1 Difficulties in Evaluation -- 5 Conclusion -- References -- Semantic Splitting of German Medical Compounds -- 1 Introduction -- 2 Related Work -- 3 Semantic Compound Splitting Approach -- 3.1 Extract Constituent Candidates from Corpus -- 3.2 Generate Corpus-Based Split Options -- 3.3 Dismiss Split Options Based on POS Tags and Suffixed -- 3.4 Generate Split Options that Include Unknown Constituents -- 3.5 Disambiguation of Split Options -- 4 Evaluation -- 4.1 Evaluation Resources -- 4.2 Evaluation Technique -- 4.3 Evaluation Results -- 4.4 Discussion and Future Work -- 5 Conclusion -- References -- A Comparison of MT Methods for Closely Related Languages: A Case Study on Czech -- Slovak and Croatian -- Slovenian Language Pairs -- 1 Introduction -- 2 State of the Art -- 3 Methodology -- 3.1 Experiment Outline -- 3.2 Test Data -- 4 Results -- 4.1 Czech -- Slovak -- 4.2 Croatian -- Slovenian -- 4.3 Inter-rater Agreement -- 5 Conclusions and Further Work -- References -- Ideas for Clustering of Similar Models of a Speaker in an Online Speaker Diarization System -- 1 Introduction -- 2 The Diarization System -- Feature Extraction and Voice Activity Detection. -- 3 Offline Clustering -- 4 Online Clustering -- 4.1 Merging Multiple GMMs into a Single One -- 4.2 Treating Multiple GMMs as Belonging to a Single Speaker -- 5 Experiments -- 6 Conclusion -- References -- Simultaneously Trained NN-Based Acoustic Model and NN-Based Feature Extractor -- 1 Introduction -- 2 Neural-Network-Based Feature Extraction -- 3 Mean Normalization, Variance Normalization and Delta Coefficients -- 4 Neural-Network-Based Acoustic Model -- 5 Experiments and Results -- 6 Conclusion and Future Work -- References -- Named Entity Recognition for Mongolian Language -- 1 Introduction -- 2 Mongolian Names -- 3 The NER System -- 3.1 Preprocessing. ; 3.2 Feature Generation.
• Sommario/riassunto: This book constitutes the refereed proceedings of the 18th International Conference on Text, Speech and Dialogue, TSD 2015, held in Pilsen, Czech Republic, in September 2015. The 67 papers presented together with 3 invited papers were carefully reviewed and selected from 138 submissions. They focus on topics such as corpora and language resources; speech recognition; tagging, classification and parsing of text and speech; speech and spoken language generation; semantic processing of text and speech; integrating applications of text and speech processing; automatic dialogue systems; as well as multimodal techniques and modelling.
• Titolo autorizzato: Text, Speech, and Dialogue
• ISBN: 3-319-24033-1
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910484934803321
• Lo trovi qui: Univ. Federico II
• Serie: Lecture Notes in Artificial Intelligence ; ; 9302