036a | XA-DE |
037b | eng |
077a | 372381340 Buchausg. u.d.T.: ‡Engelbrecht, Klaus-Peter: Estimating spoken dialog system quality with user models |
087q | 978-3-642-31590-9 |
100 | Engelbrecht, Klaus-Peter |
331 | Estimating Spoken Dialog System Quality with User Models |
410 | Berlin, Heidelberg |
412 | Springer |
425 | 2013 |
425a | 2013 |
433 | Online-Ressource (XIV, 127 p. 25 illus, digital) |
451b | T-Labs Series in Telecommunication Services |
501 | Description based upon print version of record |
517 | Estimating Spoken Dialog System Quality with User Models; Preface; Contents; Abbreviations; 1 Introduction; 1.1…Quality of Spoken Dialog Systems; 1.2…Ensuring that the System will not Fail: Usability Engineering; 1.3…Automatic Evaluation of Spoken Dialog Systems; 1.4…Evaluation Versus Learning of the ''Optimal Strategy''; 1.5…Outline and Aim of this Work; 2 MeMo: Usability Workbench; 2.1…Model Creation in the MeMo Workbench; 2.1.1 System Model; 2.1.2 User Model; 2.1.3 User Group Editor; 2.1.4 Speech Understanding Error Simulation; 2.2…Reporting; 2.2.1 Formative Usability Report. 2.2.2 Summative Report2.3…Chapter Summary; 3 Evaluation of the MeMo User Simulation---Use Case Inspire Smart Home System; 3.1…Modeling the INSPIRE Smart-Home System in MeMo; 3.2…Modeling the Experiment; 3.2.1 Rules for the Simulation; 3.3…Analysis of Simulated Corpora Using Standard Metrics; 3.3.1 Results; 3.3.1.1 High-Level Features; 3.3.1.2 Precision and Recall; 3.3.2 Comparison to Differences Between Real User Corpora; 3.3.2.1 Experiment 2 Data; 3.3.2.2 High-Level Features; 3.3.2.3 Precision and Recall; 3.4…Manual, Formative Analysis. 3.4.1 Analysis of User Actions Not Recalled by the Simulation3.4.1.1 ''Mental Model'' problem, incl. generation of new AVPs; 3.4.1.2 Flexibility in User Behavior; 3.4.1.3 Lack of Variety in Simulated User Behavior; 3.4.1.4 System Feature Not Implemented in System Model; 3.4.1.5 Experimental Artifacts; 3.4.2 Analysis of Entire Dialogs; 3.5…Discussion; 3.6…Chapter Summary; 4 Detection of Usability Problems Using an Ad-Hoc User Simulation; 4.1…Outline of the Study; 4.1.1 Experimental Data; 4.1.2 System Model; 4.1.3 User Models; 4.1.4 Speech Understanding Error Model. 7.3…An Integrated Approach. 4.2…Detection of Usability Problems4.2.1 Creating a List of Usability Problems from the Real User Data; 4.2.2 Classification and Analysis of Problems; 4.2.3 Problem Discovery in the Simulated Corpora; 4.2.4 Preparation of Data for Log File Inspection; 4.3…Discussion; 4.5…Chapter Summary; 5 Prediction of User Judgments; 5.1…Data Collection; 5.1.1 Collection of Quality Issues; 5.1.2 Selection of the System; 5.1.3 Conducting the Experiment; 5.1.4 Analysis of Judgments; 5.2…Modeling; 5.2.1 Performance Measurement; 5.2.1.1 Data Structure; 5.2.1.2 Interaction Parameters and Feature Selection. 5.2.1.3 Performance Measures5.2.2 Baseline Model with Linear Regression; 5.2.3 Modeling the Data with Markov Chains; 5.2.4 Modeling the Data with Hidden Markov Models; 5.3…Application to Data with Final Judgments; 5.3.1 Database; 5.3.2 Model Training; 5.3.3 Results; 5.4…Discussion; 5.5…Chapter Summary; 6 Application of Prediction Models in a Realistic Usage Scenario; 6.1…Outline of the Study; 6.2…Results for Real Data; 6.3…Results for Simulated Data; 6.4…Discussion; 6.5…Chapter Summary; 7 Conclusions and Future Work; 7.1…Evaluation Framework; 7.2…Modeling of User Judgments and Behavior |
527 | Buchausg. u.d.T.: ‡Engelbrecht, Klaus-Peter: Estimating spoken dialog system quality with user models |
540a | ISBN 978-3-642-31591-6 |
540a | ISBN 1-283-63142-3 ebk |
540a | ISBN 978-1-283-63142-6 MyiLibrary |
700 | |TTBM |
700 | |UYS |
700 | |TEC008000 |
700 | |COM073000 |
700b | |621.382 |
700c | |TK5102.9 |
700c | |TA1637-1638 |
700c | |TK7882.S65 |
700g | 1270821474 ST 306 |
750 | Spoken dialog systems have the potential to offer highly intuitive user interfaces, as they allow systems to be controlled using natural language. However, the complexity inherent in natural language dialogs means that careful testing of the system must be carried out from the very beginning of the design process. This book examines how user models can be used to support such early evaluations in two ways: by running simulations of dialogs, and by estimating the quality judgments of users. First, a design environment supporting the creation of dialog flows, the simulation of dialogs, and the analysis of the simulated data is proposed. How the quality of user simulations may be quantified with respect to their suitability for both formative and summative evaluation is then discussed. The remainder of the book is dedicated to the problem of predicting quality judgments of users based on interaction data. New modeling approaches are presented, which process the dialogs as sequences, and which allow knowledge about the judgment behavior of users to be incorporated into predictions. All proposed methods are validated with example evaluation studies. |
902s | 20962888X Dialogsystem |
902s | 210794070 Natürlichsprachiges System |
902s | 208934502 Gesprochene Sprache |
902s | 211441554 Benutzermodell |
902s | 209907495 Leistungsbewertung |
902s | 208861580 Benutzerfreundlichkeit |
902s | 213452766 Sprachqualität |
902s | 209112050 Simulation |
902s | 211692794 Testen |
902s | 209555157 Benutzerverhalten |
012 | 373431686 |
081 | Engelbrecht, Klaus-Peter: Estimating Spoken Dialog System Quality with User Models |
100 | Springer E-Book |
125a | Elektronischer Volltext - Campuslizenz |
655e | $uhttp://dx.doi.org/10.1007/978-3-642-31591-6 |