Graph Learning and Network Science for Natural Language Processing

Cover
Half Title
Series Page
Title Page
Copyright Page
Table of Contents
Editors
Contributors
Preface
Chapter 1 Graph of Words Model for Natural Language Processing
1.1 Introduction
1.1.1 Lexical and Morphological Analysis
1.1.2 Syntactic Analysis
1.1.3 Semantic Analysis
1.1.4 Discourse Integration
1.1.5 Pragmatic Analysis
1.2 Machine Learning and Text Modelling
1.3 BoW Model
1.3.1 Introduction
1.3.1.1 Step 1: Collect the Data
1.3.1.2 Step 2: Vocabulary Design
1.3.1.3 Step 3: Document Vectors Creation
1.3.1.4 Scoring Words
1.3.2 Limitations of the BoW Model
1.4 Graph of Words (GoW) Model
1.4.1 Basic Terminology of Graphs
1.4.1.1 Real-world Graphs
1.4.1.2 Graphs in Linguistics
1.4.2 Semantic Similarity and Ambiguity
1.4.3 How to Build a GoW
1.4.3.1 Preliminary Concepts
1.4.4 Construction of a GoW
1.4.5 Use of GoW in Text Mining
1.4.6 GoW Mining
1.4.6.1 Graph Degeneracy
1.4.6.2 K-core Decomposition
1.4.6.3 K-truss
1.5 Discussion and Future Scope
References
Chapter 2 Application of NLP Using Graph Approaches
2.1 Introduction
2.1.1 What Is a Graph?
2.2 Graph Embeddings
2.3 Dynamic Graph of Words
2.4 Cross-lingual and Multilingual Graphical Approaches
2.5 Topological Analysis of Graphs
2.6 Adversarial Networks for Natural Language Processing
2.7 Heterogeneous Information Networks for Textual Information
2.8 Summary of Ontology and Knowledge Graphs
2.9 Topic Identification
2.10 Major Processes of NLP Using Graphical Approaches and Their Applications in the Real World
2.10.1 Summarization
2.10.1.1 News
2.10.1.2 Assignments and E-learning
2.10.1.3 Summarization of Financial or Legal Documents
2.10.2 Semi-supervised Passage Retrieval
2.10.3 Keyword Extraction
2.10.3.1 The Steps of the TextRank Algorithm
2.10.4 Information Extraction
2.10.5 Question Answering
2.10.6 Cross-language Information Retrieval
2.10.7 Term Weighting
2.10.8 Topic Segmentation
2.10.8.1 Graph-based Topic Segmentation
2.10.9 Machine Translation
2.10.9.1 Graph-based Machine Translation
2.10.10 Discourse Analysis
2.11 Conclusion and Future Scope of NLP
2.12 Datasets for NLP Applications
References
Chapter 3 Graph-based Extractive Approach for English and Hindi Text Summarization
3.1 Introduction
3.2 Text Summarization Approaches
3.2.1 Text Summarization Based on Number of Documents
3.2.2 Text Summarization Based on the Summary’s Purpose
3.2.3 Text Summarization Techniques
3.2.4 Text Summarization Based on Level of Language
3.2.5 Text Summarization Based on Output Style
3.2.6 Text Summarization Based on the Summary’s Characteristics
3.3 Literature Survey
3.4 Graph-based Algorithms
3.4.1 PageRank Algorithm
3.4.2 Text Rank Algorithm
3.5 TF-IDF Algorithm
3.6 Methodology
3.7 Experimental Results
3.7.1 English Original Text
3.7.2 Summary Produced Using the Text Rank Algorithm
3.7.3 Summary Produced Using the TF-IDF Algorithm
3.7.4 Hindi Original Text
3.7.5 Summary Produced Using the Text Rank Algorithm
3.7.6 Summary Produced Using the TF-IDF Algorithm
3.8 Conclusions and Future Directions
References
Chapter 4 Graph Embeddings for Natural Language Processing
4.1 Introduction
4.1.1 Natural Language and Natural Language Processing
4.1.2 Processing: A Module in Machine Learning
4.2 Computational Techniques
4.2.1 How NLP Works
4.2.2 Graph Embeddings
4.2.2.1 Graph
4.2.2.2 Embedding
4.2.2.3 Graph Embeddings
4.2.2.4 Word Embeddings: Classic Example
4.3 The Singular Value Decomposition for “Graph Embeddings”
4.4 Predictive Methods
4.4.1 Word2vec
4.4.1.1 CBOW: Continuous Bag of Words [5]
4.4.1.2 Skip-gram Model
4.5 More Embedding Techniques
4.6 Conclusion
4.7 Case Study: Neo4j Lab Implementations
References
Chapter 5 Natural Language Processing with Graph and Machine Learning Algorithms-based Large-scale Text Document Summarization and Its Applications
5.1 Introduction
5.1.1 Types of Text Summary
5.2 Text Summarization and Machine Learning
5.2.1 What Is Graph ML?
5.2.2 Simple Algorithmic Approach
5.3 Literature Survey
5.3.1 Gap Analysis
5.4 Problem Statement
5.5 System Architecture
5.6 Graph-based Solutions
5.7 Conclusion
References
Chapter 6 Ontology and Knowledge Graphs for Semantic Analysis in Natural Language Processing
6.1 Introduction
6.2 Background
6.2.1 Semantics in NLP
6.2.1.1 Meaning Representation
6.2.2 Semantic Analysis in Natural Language Processing
6.3 Semantic Technologies
6.3.1 Ontology Essentials
6.3.2 Formalization of an Ontology
6.3.3 Description Logics
6.3.4 Ontological Languages
6.3.5 Knowledge Graphs
6.3.5.1 Ontology and Knowledge Graphs
6.3.5.2 Property Graphs
6.3.5.3 Comparison of KGs, PGs, and Ontologies
6.4 The Role of Ontology and Knowledge Graphs in Semantic Analysis
6.4.1 The Role of Semantic Technology in Knowledge-based NLP Applications
6.4.2 The Role of Semantic Technology Machine Learning NLP Applications
6.4.3 The Role of Natural Language Processing in Ontology Generation
6.5 Review of Developments in Ontological Semantic Analysis
6.6 Summary
References
Chapter 7 Ontology and Knowledge Graphs for Natural Language Processing
7.1 Introduction
7.1.1 Ontology
7.1.1.1 The Core Ideas of Ontology
7.2 Natural Language Processing
7.2.1 Dealing with Huge Amounts of Unstructured Data
7.2.2 Structuring Data to Support Intelligent Systems
7.2.3 Challenges with NLP
7.3 Ontology and Knowledge Graphs for NLP
7.3.1 Ontology
7.3.2 Knowledge Graphs
7.4 Ontological Languages
7.5 Conclusion
References
Chapter 8 Perfect Coloring by HB Color Matrix Algorithm Method
8.1 Introduction
8.2 Preliminaries
8.2.1 HB Color Matrix
8.2.2 Perfect Coloring
8.3 Results
8.3.1 Perfect HB Color Matrix
8.3.1.1 Example of a PHBCM
8.3.1.2 Properties of a PHBCM
8.3.2 Algorithm of Perfect Coloring by HB Color Matrix Method
8.4 Illustration of the Perfect HB Color Matrix Method
8.5 Python Program for Graph Coloring by PHBCM
8.6 The Perfect Chromatic Numbers for Some Standard Graphs Using the PHBCM Algorithm Method
8.7 Conclusion
References
Chapter 9 Cross-lingual Word Sense Disambiguation Using Multilingual Co-occurrence Graphs
9.1 Introduction
9.2 Evaluation of WSD
9.2.1 Types of WSD
9.2.2 Difficulties in Word Sense Disambiguation
9.2.2.1 Differences between Dictionaries
9.2.2.2 Part of Speech Tagging
9.2.2.3 Inter-judge Variance
9.2.2.4 Pragmatics
9.2.2.5 Sense Inventories and Task-dependent Algorithms
9.2.2.6 Sense Discreteness
9.3 Approaches to Word Sense Disambiguation
9.3.1 Dictionary and Knowledge-based Methods
9.3.2 Supervised Methods
9.3.3 Semi-Supervised Methods
9.3.4 Unsupervised Methods
9.4 Graph-based Cross-Lingual Word Sense Disambiguation
9.4.1 MultiMirror Model
9.4.2 UHD Model
9.4.3 Co-occurrence Graphs for WSD in the Biomedical Domain
9.4.4 WSD Based on Word Similarity Calculation Using Weighted Voronoi Regions from a Knowledge Graph
9.4.5 Graph Convolutional Networks for WSD
9.4.6 The Context Expansion Approach in Graph WSD
9.4.7 WSD Using WordNet Knowledge Graphs
9.5 Applications of WSD
9.6 Conclusions and Future Scope
References
Chapter 10 Study of Current Learning Techniques for Natural Language Processing for Early Detection of Lung Cancer
10.1 Introduction
10.2 Rationale and Significance of the Study
10.3 Motivation
10.4 Learning Techniques
10.4.1 Data Annotation
10.4.2 Word Embedding
10.4.3 NER Process
10.4.4 Relation Classification Process
10.4.5 Prediction Performance Step
10.5 Related Work
10.6 Discussion
10.7 Conclusion
References
Chapter 11 A Critical Analysis of Graph Topologies for Natural Language Processing and Their Applications
11.1 Introduction
11.1.1 Natural Language Processing
11.1.2 Tools and Libraries for NLP
11.1.3 Graph of Words and Graph -based Natural Language Generation
11.4 Graph Embedding in NLP
11.4.1 Node Embeddings
11.4.1.1 Matrix Factorization Methods
11.4.1.2 Graph Neural Network Methods
11.4.1.3 Random Walk (RW) Methods
11.4.1.4 Applications of Node Embedding
11.4.2 Relation Embedding
11.4.2.1 Knowledge-based Relation Embedding
11.4.2.2 Unsupervised Relation Embedding
11.4.2.3 Applications of Relation Embedding
11.5 Graph Topologies for NLP Applications
11.5.1 Critical Analysis of Graph Architectures for NLP Applications
11.5.1.1 Text Formation, Conversation, and Generation
11.5.1.2 Language Rules and Classification
11.5.1.3 Context
11.5.1.4 Machine Translation
11.5.1.5 Knowledge Mining and Demonstration
11.6 Conclusion and Future Work
References
Chapter 12 Graph-based Text Document Extractive Summarization
12.1 Introduction
12.2 Extractive Summarization
12.2.1 Commonly used Features in Extractive Summarization Method
12.2.2 Summary Length
12.3 Graph-based Methods for Extractive Summarization
12.3.1 Graph-based Ranking Algorithm
12.3.2 Weighted/unweighted Simple Graph
12.3.3 Heterogeneous Graph Model
12.3.4 Correlation Graph Model
12.3.5 Semantic Graph Model
12.3.6 Hypergraph Model
12.3.6.1 Hypergraph Construction
12.3.7 Semigraph Model
12.4 Conclusion
References
Chapter 13 Applications of Graphical Natural Language Processing
13.1 Graph Theory in Natural Language Processing
13.2 Text Summarization
13.3 Keyword Extraction
13.4 Graph -oriented Topic Analysis
13.5 Topic Segmentation
13.6 Discourse Relationships
13.7 Machine Translation
13.8 Multilingual Retrieval of Information Based on Graphs
13.9 Information Retrieval Using Graphs
13.10 Graph -based Question Answering
References
Chapter 14 Analysis of Medical Images Using Machine Learning Techniques
14.1 Introduction
14.1.1 Overview
14.1.2 Motivation
14.1.3 Objective
14.1.3.1 Study of Different Segmentation Techniques
14.1.3.2 Selection for Appropriate Features
14.1.3.3 Performance and Classification of the Proposed Model
14.1.4 Expected Outcomes
14.2 Literature Survey
14.3 The Problem Domain and Proposed Solution
14.3.1 Domain Description
14.3.2 Problem domain
14.3.3 Solution Domain
14.3.4 Algorithms Used in the Study
14.3.4.1 Otsu’s Method
14.3.4.2 Wavelet Transformation
14.3.4.3 Principal Component Analysis
14.3.4.4 Gray Level Co-occurrence Matrix
14.3.4.5 Support Vector Machine
14.3.5 Our Proposed Algorithm
14.4 Implementation
14.4.1 Tools and Techniques
14.4.2 MATLAB
14.4.3 Methodology
14.5 Result Analysis
14.5.1 Means
14.5.2 Standard Deviation
14.5.3 Entropy
14.5.4 RMS
14.5.5 Variance
14.5.6 Smoothness
14.5.7 Kurtosis
14.5.8 Skewness
14.5.9 IDM
14.5.10 Contrast
14.5.11 Correlation
14.5.12 Energy
14.5.13 Homogeneity
14.5.14 Kernel Functions
14.6 Conclusion and Future Work
14.6.1 Conclusion
14.6.2 Future Work
References
Index