Natural language response generation in mixed-initiative dialogs.

Yip Wing Lin Winnie.Thesis (M.Phil.)--Chinese University of Hong Kong, 2004.Includes bibliographical references (leaves 102-105).Abstracts in English and Chinese.Chapter 1 --- Introduction --- p.1Chapter 1.1 --- Overview --- p.1Chapter 1.2 --- Thesis Goals --- p.3Chapter 1.3 --- Thesis Outline --- p.5Chapter 2 --- Background --- p.6Chapter 2.1 --- Natural Language Generation --- p.6Chapter 2.1.1 --- Template-based Approach --- p.7Chapter 2.1.2 --- Rule-based Approach --- p.8Chapter 2.1.3 --- Statistical Approach --- p.9Chapter 2.1.4 --- Hybrid Approach --- p.10Chapter 2.1.5 --- Machine Learning Approach --- p.11Chapter 2.2 --- Evaluation Method --- p.12Chapter 2.2.1 --- Cooperative Principles --- p.13Chapter 2.3 --- Chapter Summary --- p.13Chapter 3 --- Natural Language Understanding --- p.14Chapter 3.1 --- The CUHK Restaurant Domain --- p.15Chapter 3.2 --- "Task Goals, Dialog Acts, Concept Categories and Annotation" --- p.17Chapter 3.2.1 --- Task Goals (TGs) and Dialog Acts (DAs) --- p.17Chapter 3.2.2 --- Concept Categories (CTG/CDA) --- p.20Chapter 3.2.3 --- Utterance Segmentation and Annotation --- p.21Chapter 3.3 --- Task Goal and Dialog Act Identification --- p.22Chapter 3.3.1 --- Belief Networks Development --- p.22Chapter 3.3.2 --- Task Goal and Dialog Act Inference --- p.24Chapter 3.3.3 --- Network Dimensions --- p.25Chapter 3.4 --- Chapter Summary --- p.29Chapter 4 --- Automatic Utterance Segmentation --- p.30Chapter 4.1 --- Utterance Definition --- p.31Chapter 4.2 --- Segmentation Procedure --- p.33Chapter 4.2.1 --- Tokenization --- p.35Chapter 4.2.2 --- POS Tagging --- p.36Chapter 4.2.3 --- Multi-Parser Architecture (MPA) Language Parsing --- p.38Chapter 4.2.4 --- Top-down Generalized Representation --- p.40Chapter 4.3 --- Evaluation --- p.47Chapter 4.3.1 --- Results --- p.47Chapter 4.3.2 --- Analysis --- p.48Chapter 4.4 --- Chapter Summary --- p.50Chapter 5 --- Natural Language Response Generation --- p.52Chapter 5.1 --- System Overview --- p.52Chapter 5.2 --- Corpus-derived Dialog State Transition Rules --- p.55Chapter 5.3 --- Hand-designed Text Generation Templates --- p.56Chapter 5.4 --- Performance Evaluation --- p.59Chapter 5.4.1 --- Task Completion Rate --- p.61Chapter 5.4.2 --- Grice's Maxims and Perceived User Satisfaction --- p.62Chapter 5.4.3 --- Error Analysis --- p.64Chapter 5.5 --- Chapter Summary --- p.65Chapter 6 --- Bilingual Response Generation using Semi-Automatically- Induced Response Templates --- p.67Chapter 6.1 --- Response Data --- p.68Chapter 6.2 --- Semi-Automatic Grammar Induction --- p.69Chapter 6.2.1 --- Agglomerative Clustering --- p.69Chapter 6.2.2 --- Parameters Selection --- p.70Chapter 6.3 --- Application to Response Grammar Induction --- p.71Chapter 6.3.1 --- Parameters Selection --- p.73Chapter 6.3.2 --- Unsupervised Grammar Induction --- p.76Chapter 6.3.3 --- Post-processing --- p.80Chapter 6.3.4 --- Prior Knowledge Injection --- p.82Chapter 6.4 --- Response Templates Generation --- p.84Chapter 6.4.1 --- Induced Response Grammar --- p.84Chapter 6.4.2 --- Template Formation --- p.84Chapter 6.4.3 --- Bilingual Response Templates --- p.89Chapter 6.5 --- Evaluation --- p.89Chapter 6.5.1 --- "Task Completion Rate, Grice's Maxims and User Sat- isfaction" --- p.91Chapter 6.6 --- Chapter Summary --- p.94Chapter 7 --- Conclusion --- p.96Chapter 7.1 --- Summary --- p.96Chapter 7.2 --- Contributions --- p.98Chapter 7.3 --- Future Work --- p.100Bibliography --- p.102Chapter A --- Domain-Specific Task Goals in the CUHK Restaurants Do- main --- p.107Chapter B --- Full List of VERBMOBIL-2 Dialog Acts --- p.109Chapter C --- Dialog Acts for Customer Requests and Waiter Responsesin the CUHK Restaurants Domain --- p.111Chapter D --- Grammar for Task Goal and Dialog Act Identification --- p.116Chapter E --- Utterance Definition --- p.119Chapter F --- Dialog State Transition Rules --- p.121Chapter G --- Full List of Templates Selection Conditions --- p.125Chapter H --- Hand-designed Text Generation Templates --- p.130Chapter I --- Evaluation Test Questionnaire for Dialog System in the CUHK Restaurant Domain --- p.135Chapter J --- POS Tags --- p.137Chapter K --- Full List of Lexicon and contextual rule modifications --- p.139Chapter L --- Top-down Generalized Representations --- p.141Chapter M --- Sample Outputs for Automatic Utterance Segmentation --- p.144Chapter N --- Induced Grammar --- p.145Chapter O --- Seeded Categories --- p.148Chapter P --- Semi-Automatically-Induced Response Templates --- p.150Chapter Q --- Details of the Statistical Testing Regarding Grice's Maxims and User Satisfaction --- p.15