Text Analytics: Advances and Challenges

Preface
Contents
Techniques, Methods and Models
Text Analytics: Present, Past and Future
	1 Introduction
	2 The Text Analysis Pioneers
	3 Timeline Representation Framework
	4 Text Analytics Process
	5 Scientific Research and Collaborations
	References
Unsupervised Analytic Strategies to Explore Large Document Collections
	1 Introduction
	2 From Unstructured to Structured Data
		2.1 Data Acquisition and Pre-treatment
		2.2 Data Representation and Organisation
	3 Looking for Knowledge: A Comparison Between Text Analytics
		3.1 Factorial-Based Approach
		3.2 Network-Based Approach
		3.3 Probabilistic Approach
	4 Discussion
	References
Studying Narrative Flows by Text Analysis and Network Text Analysis
	1 Introduction
	2 A Joint Approach: Method and Procedures
	3 Young People’s Job Perceptions: A Case Study
	4 Results and Discussion
	5 Conclusions
	References
Key Passages : From Statistics  to Deep Learning
	1 Introduction
	2 Key Passages and Statistics
		2.1 Definition
		2.2 Implementations
		2.3 Examples
	3 Key Passages and Deep Learning
		3.1 Most Relevant Passages Detected by the Neural Network
		3.2 Learning Linguistic Features
		3.3 Examples
	4 Conclusion
		4.1 Discussion
		4.2 Perspectives
	References
Concentration Indices for Dialogue Dominance Phenomena in TV Series: The Case of the Big Bang Theory
	1 Introduction
	2 The Dataset
	3 Representation of Dialogues
	4 Dominance in Dialogues
	5 Conclusions
	References
A Conversation Analysis of Interactions in Personal Finance Forums
	1 Introduction
	2 Datasets
	3 Dominance in a Thread
	4 Interactions Between Post Submitters
	5 Conclusions
	References
Dictionaries and Specific Languages
Big Corpora and Text Clustering: The Italian Accounting Jurisdiction Case
	1 Introduction
	2 Data and Method
	3 The Main Results
		3.1 Preprocessing and Building of a Dictionary
		3.2 Visualization and Clustering
	4 Conclusions and the Future Steps
	References
Lexicometric Paradoxes of Frequency: Comparing VoBIS and NVdB
	1 Introduction
	2 Scholastic Italian
	3 Methodological Aspects
	4 Base Vocabularies Compared: Frequencies in the NVdB and VoBIS
	5 Conclusions and Future Prospects
	References
Emotions and Dense Words in Emotional Text Analysis: An Invariant or a Contextual Relationship?
	1 Emotional Text Analysis: Specific Characteristics and Issues in Research
	2 The Research Project
		2.1 Testing the Consistency of Dense and Non-dense Words: Data Construction and Methodology
		2.2 Study of the Relationship Between the Range of Word Consistency and the Groups of Dictionaries
		2.3 Definition of a Basic List of Dense and Non-dense Words
	3 Conclusion
	References
Text Mining of Public Administration Documents: Preliminary Results on Judgments
	1 Introduction and Objectives
	2 Methods
	3 Data and Results
		3.1 A Document Warehouse of the Judgments
		3.2 Companies’ Text Mining
		3.3 The Record Linkage Phase
	4 Conclusions and Further Steps
	Annex
	References
Using the First Axis of a Correspondence Analysis as an Analytic Tool
	1 Research Question and Scientific Background
	2 Methodology and Preparation of Textual Data
	3 Linguistic and Methodological Results from Correspondence Analysis
	4 Discussion: Can the First Axis Match Any Data Feature, Is Orality Feature Really the Main One?
		4.1 Challenging with Another Bipolar Textual Feature: The Verse/Prose Opposition
		4.2 Challenging with the Most Fine-Grained Units: Individual Texts
		4.3 Challenging with a Linear Feature: Time
		4.4 Corpus Scale Effect
	5 Complementary Tools to Analyze 1D Gradient in Textual Data
		5.1 Gradient-Ordered Barplot
		5.2 Back-to-Text Close Reading by Getting Representative Words for Each Side of the First Axis
	6 Conclusion
	References
Discursive Functions of French Modal Forms: What Can Correspondence Analysis Tell Us About Genre and Diachronic Variation?
	1 Introduction
	2 Analysis of Combinatory Profiles of Epistemic Adverbs (E-ADVs)
	3 Correspondence Analysis (CA) of E-ADVs
	4 Correspondence Analysis (CA) of Necessity and Possibility Modal Forms
	5 Perspectives
	References
Multilingual Text Analysis
How to Think About Finding a Sign for a Multilingual and Multimodal French-Written/French Sign Language Platform?
	1 Introduction
	2 Confrontation of Two Paradigms
		2.1 A Phonological Hypothesis of SL
		2.2 A Morphemic Hypothesis of SL
	3 Significant Elements in the Search for a Sign for a Written French Multilingual and Multimodal Platform/LSF
		3.1 Context of a Sign Search in a Bilingual Corpus Written Language/SL
		3.2 Automation of Indexing
		3.3 Current Situation and Limit
		3.4 Description and Search Criteria
	4 Conclusion
	References
Corpus in “Natural” Language Versus “Translation” Language: LBC Corpora, A Tool for Bilingual Lexicographic Writing
	1 Introduction
	2 The LBC Corpora
	3 Comparison Between “Natural Language” and “Translation” Corpus: A Perspective Halfway Between Descriptive and Prescriptive Translatology
		3.1 Descriptive and Prescriptive Information in LBC Dictionaries: Universals and Deviations
	4 Natural Language Versus Translated Language: Observation of the Corpus
		4.1 Zero Frequency in Natural Language Texts
		4.2 Frequency Difference in Source Texts Compared to Target Texts
	5 Conclusion
	References
The Conditional Perfect, A Quantitative Analysis in English-French Comparable-Parallel Corpora
	1 Introduction
	2 Methods
	3 Results and Analysis
	4 Discussion
	5 Conclusion
	References
Repeated and Anaphoric Segments Applied to Trilingual Knowledge Extraction
	1 Introduction
	2 Definitions and Conceptual Scope
	3 Corpus
	4 Repetitions in ENRGFR and ENRGUS
	5 English–Chinese Parallel Repeated Segments
	6 Recurrences in Chinese Through the Prism of Their Poly-cooccurrences Experience
	7 Conclusion
	References
Applications
Looking for Topics: A Brief Review
	1 Introduction
	2 An Outline of the Contents of Shakespeare’s Sonnets
		2.1 Theme, Topic, Motif
		2.2 Eight Themes Derived from Expert Commentaries
	3 Six Selected Methods for Topic Research
	4 Excerpts from the List of 49 Topics (Limited to Two Topics Per Method)
		4.1 Rotated Factor Analysis (Rotation Oblimin) (2 Topics Out of 6)
		4.2 Fragmented Correspondence Analysis (FCA) (2 Topics Out of 7)
		4.3 Logarithmic Analysis (Spectral Mapping) (2 Topics Out of 8)
		4.4 Latent Semantic Analysis (2 Topics Out of 8)
		4.5 NMF Topics (2 Topics Out of 10)
		4.6 Latent Dirichlet Allocation LDA (2 Topics Out of 10)
	5 A Synthesis of Produced Topics
	References
Where Are the Social Sciences Going to? The Case of the EU-Funded SSH Research Projects
	1 Introduction: The EU “Clouded-Sky” Research Funding and Its Role in Affecting SSH Thematic Foci
	2 The Socio-political Dimensions of Scientific Research
	3 The Corpus and Methodology
	4 The Main Findings
	5 Conclusion: The SSH and the “European Spirit”
	References
Topic Modeling of Twitter Conversations: The Case of the National University of Colombia
	1 Introduction
	2 Method
		2.1 Data Collection and Pre-processing
		2.2 Statistical Analysis
	3 Results
		3.1 Protest Topics
		3.2 Anti-reform Topics
		3.3 Investment Topics
		3.4 Peace Topic
	4 Conclusions
	References
Analyzing Occupational Safety Culture Through Mass Media Monitoring
	1 Introduction
	2 The Database
		2.1 Portrayal of the Tool
	3 Methodology and Data
	4 Main Results and Discussion
	5 Conclusions
	References
What Volunteers Do? A Textual Analysis of Voluntary Activities in the Italian Context
	1 Introduction
	2 Volunteering Context in Italy
	3 Data and Statistical Approach
	4 Main Findings of the Analysis
	5 Discussion and Conclusion
	References
Free Text Analysis in Electronic Clinical Documentation
	1 Introduction
	2 Materials and Methods
		2.1 Corpus
		2.2 Linguistic Filtering Procedure
		2.3 Statistical Analysis
	3 Results
	4 Conclusions
	References
Educational Culture and Job Market: A Text Mining Approach
	1 Introduction
	2 The Emotional Text Mining
	3 Methods
	4 Main Results and Discussion
	5 Conclusion
	References
Author Index
Subject Index