Trích xuất cụm từ tiếng Trung tự động - Luận án tiến sĩ của Xu Ruifeng

Trường ĐH

The Hong Kong Polytechnic University

Chuyên ngành

Computing

Tác giả

Ẩn danh

Thể loại

Luận án tiến sĩ

Năm xuất bản

Số trang

214

Thời gian đọc

33 phút

Lượt xem

0

Lượt tải

0

Phí lưu trữ

50 Point

Mục lục chi tiết

Certificate of Originality

Publications Arising from the Thesis

List Of Figures

1. Chapter 1 Introduction

1.1. Basic Concepts and Thesis Scope

1.2. Motivation and Problem Statement

1.3. Research Objectives and Thesis Scope

2. Chapter 2: Literature Review

2.1. Review of Automatic Collocation Extraction Techniques

2.1.1. Window-based Statistical Collocation Extraction Approach

2.1.2. Syntax-based Collocation Extraction Approach

2.1.3. Collocation Extraction using Semantic Information

2.2. Review of Automatic Shallow Parsing

2.2.1. Statistic-based Shallow Parsing

2.2.2. Rule-based Shallow Parsing

3. Chapter 3 Collocation Extraction Based on Lexical Statistics

4. Chapter 4 Collocation Extraction Based on Lexical Statistics

4.1. Preparation of Training Corpus and Answer Set

4.2. Applying Xtract to Chinese Collocation Extraction: CXtract

4.3. Improving CXtract: CXtractII

4.4. A New Collocation Extraction System: CXtract2

4.4.1. The Framework Design

4.4.2. Construct a Word Co-occurrence Database for CXtract2

4.5. Evaluation Of CXtract2

4.6. Evaluations of Statistical Collocation Extraction Algorithms

5. Chapter 5 Multi-Stage Collocation Extraction

5.1. Categorization of Chinese Collocations

5.2. Characteristic Analysis of Typical Collocations

5.3. The Design of A Multi-Stage Collocation Extraction System

5.3.1. Additional Feature Selections

5.3.2. Applying the Heuristic Rules to Eliminate Pseudo Collocations

5.3.3. The New Multi-stage Extraction Algorithm

5.3.4. Parameter Optimization based on Perceptron Training Rule

5.4. Experimental Results and Evaluations

5.4.1. Experimental Data Preparation

5.4.2. Experiments on Type 1 and Type 2 Collocation Extraction in Stage 3

5.4.3. Experiments on Weight Parameter Optimization

5.4.4. Experiments on Multi-stage Collocation Extraction of Stage 1-3

5.4.5. Experiments on Pseudo collocation Filtering by Using Heuristic Rules

5.4.6. Experiments on Evaluating the Complete Collocation Extraction System

5.5. Chapter Summarization

6. Chapter 6 The Design and Development of Chinese Shallow Treebank and Automatic Chunkers

6.1. The Design and Development of PolyU Treebank

6.1.1. Basic Concepts and Background of Shallow Treebank

6.3. Annotation Guideline Design

6.4. Implementation of the PolyU Treebank

6.5. Quality Assurance and Annotation Process

6.6. Contributions of PolyU Treebank

6.7. The Design and Development of Automatic Chunkers

6.7.1. Chunking Scope and Representation

6.7.2. Chunking with POS Features

6.7.3. Chunking with Lexicalized Features

6.7.4. Experiments and Evaluations

7. Chapter 7 Collocation Extraction Using Chunking Information

7.1. Syntactic Representation and Collocation Patterns Extraction

7.1.1. Syntactic Representation

7.2. Support Collocation Patterns Extraction

7.3. Reject Collocation Patterns Extraction

7.4. Incorporating Syntactic Patterns into Collocation Extraction

7.5. Experimental Results and Evaluations

7.6. An Overall Comparison

8. Chapter 8 Applying Collocations for Handwritten Character Recognition

8.1. Post-processing Techniques for Improving HCCR Systems

8.2. Applying Collocation Database in Post-processing Systems

8.3. Experimental Results

8.4. Chapter Summarization

9. Chapter 9 Conclusion and Future Work

Appendix 1 The POS Tag Set

Appendix 2 The DTD (Document Type Definition) File Content

Appendix 3 Examples of An Shallow Annotated Text

Appendix 4 Examples of Collocations

Appendix 5 Examples of Heuristic Rules for Pseudo Collocation Filtering

Xem trước tài liệu
Tải đầy đủ để xem toàn bộ nội dung
Luận án tiến sĩ: The study on automatic Chinese collocation extraction

Tải xuống file đầy đủ để xem toàn bộ nội dung

Tải đầy đủ (214 trang)

Từ khóa và chủ đề nghiên cứu


Câu hỏi thường gặp

Luận án liên quan

Chia sẻ tài liệu: Facebook Twitter