Text Mining for Qualitative Data Analysis in the Social Sciences (eBook)

A Study on Democratic Discourse in Germany

Gregor Wiedemann (Autor)

eBook Download: PDF

2016 | 1st ed. 2016
XVII, 294 Seiten
Springer Fachmedien Wiesbaden GmbH (Verlag)
9783658153090 (ISBN)

Infos

Gregor Wiedemann evaluates text mining applications for social science studies with respect to conceptual integration of consciously selected methods, systematic optimization of algorithms and workflows, and methodological reflections relating to empirical research. In an exemplary study, he introduces workflows to analyze a corpus of around 600,000 newspaper articles on the subject of 'democratic demarcation' in Germany. He provides a valuable resource for innovative measures to social scientists and computer scientists in the field of applied natural language processing. 

Gregor Wiedemann holds a PhD degree from Leipzig University, Germany. He is the coordinator of the discipline-specific working groups in the CLARIN-D project, which develops a European virtual research infrastructure for digital language data analysis in the social sciences and humanities.

Autor

Gregor Wiedemann holds a doctoral degree from Leipzig University, Germany. He is the coordinator of the discipline-specific working groups in the CLARIN-D project, which develops a European virtual research infrastructure for digital language data analysis in the social sciences and humanities.

Inhalt

Preface 6
Contents 7
List of Figures 10
List of Tables 11
List of Abbreviations 13
1. Introduction: Qualitative DataAnalysis in a Digital World 16
1.1. The Emergence of “Digital Humanities” 18
1.2. Digital Text and Social Science Research 23
1.3. Example Study: Research Question andData Set 26
1.3.1. Democratic Demarcation 27
1.3.2. Data Set 27
1.4. Contributions and Structure of the Study 29
2. Computer-Assisted TextAnalysis in the Social Sciences 32
2.1. Text as Data between Quality and Quantity 32
2.2. Text as Data for Natural Language Processing 37
2.2.1. Modeling Semantics 37
2.2.2. Linguistic Preprocessing 41
2.2.3. Text Mining Applications 43
2.3. Types of Computational Qualitative DataAnalysis 49
2.3.1. Computational Content Analysis 55
2.3.2. Computer-Assisted Qualitative Data Analysis 58
2.3.3. Lexicometrics for Corpus Exploration 60
2.3.4. Machine Learning 64
3. Integrating Text Mining Applications for ComplexAnalysis 70
3.1. Document Retrieval 71
3.1.1. Requirements 71
3.1.2. Key Term Extraction 74
3.1.3. Retrieval with Dictionaries 81
3.1.4. Contextualizing Dictionaries 84
3.1.5. Scoring Co-Occurrences 86
3.1.6. Evaluation 89
3.1.7. Summary of Lessons Learned 97
3.2. Corpus Exploration 99
3.2.1. Requirements 100
3.2.2. Identification and Evaluation of Topics 103
3.2.3. Clustering of Time Periods 115
3.2.4. Selection of Topics 120
3.2.5. Term Co-Occurrences 123
3.2.6. Keyness of Terms 127
3.2.7. Sentiments of Key Terms 127
3.2.8. Semantically Enriched Co-Occurrence Graphs 130
3.2.9. Summary of Lessons Learned 137
3.3. Classification for Qualitative Data Analysis 140
3.3.1. Requirements 143
3.3.2. Experimental Data 147
3.3.3. Individual Classification 150
3.3.4. Training Set Size and Semantic Smoothing 155
3.3.5. Classification for Proportion and Trend Analysis 161
3.3.6. Active Learning 170
3.3.7. Summary of Lessons Learned 180
4. Exemplary Study: DemocraticDemarcation in Germany 182
4.1. Democratic Demarcation 182
4.2. Exploration 189
4.2.1. Democratic Demarcation from 1950–1956 190
4.2.2. Democratic Demarcation from 1957–1970 193
4.2.3. Democratic Demarcation from 1971–1988 195
4.2.4. Democratic Demarcation from 1989–2000 198
4.2.5. Democratic Demarcation from 2001–2011 200
4.3. Classification of Demarcation Statements 202
4.3.1. Category System 203
4.3.2. Supervised Active Learning of Categories 207
4.3.3. Category Trends and Co-Occurrences 210
4.4. Conclusions and Further Analyses 224
5. V-TM – A MethodologicalFramework for Social Science 227
5.1. Requirements 230
5.1.1. Data Management 233
5.1.2. Goals of Analysis 234
5.2. Workflow Design 237
5.2.1. Overview 238
5.2.2. Workflows 242
5.3. Result Integration and Documentation 252
5.3.1. Integration 253
5.3.2. Documentation 255
5.4. Methodological Integration 257
6. Summary: Integrating Qualitative and ComputationalText Analysis 264
6.1. Meeting Requirements 265
6.2. Exemplary Study 268
6.3. Methodological Systematization 269
6.4. Further Developments 270
A. Data Tables, Graphs andAlgorithms 274
Bibliography 284

Produkt-Details

• Erscheint lt. Verlag: 23.8.2016
• Reihe/Serie: Kritische Studien zur Demokratie
• Reihe/Serie: Kritische Studien zur Demokratie
• Zusatzinfo: XVII, 294 p. 24 illus.
• Verlagsort: Wiesbaden
• Sprache: englisch
• Themenwelt: Sozialwissenschaften ► Politik / Verwaltung
• Themenwelt: Sozialwissenschaften ► Soziologie ► Empirische Sozialforschung
• Schlagworte: Big Data • Democracy • Modeling Semantics • Qualitative Text Analyses • Text Qualification • Workflows
• ISBN-13: 9783658153090 / 9783658153090