Computational Intelligence Methods for Sentiment Analysis in Natural Language Processing Applications

Computational Intelligence Methods for Sentiment Analysis in Natural Language Processing Applications

Sentiment Analysis has become increasingly important in recent years for nearly all online applications. Sentiment Analysis depends heavily on Artificial Intelligence (AI) technology wherein computational intelligence approaches aid in deriving the opinions/emotions of human beings. With the vast in...

Descripción completa

Detalles Bibliográficos
Otros Autores: Hemanth, D. Jude, editor (editor)
Formato: Libro electrónico
Idioma:Inglés
Publicado: Cambridge, MA : Mara Conner [2024]
Edición:First edition
Materias:
Computational intelligence.
Natural language processing (Computer science) > Computer programs.
Artificial intelligence > Computer programs.
Ver en Biblioteca Universitat Ramon Llull:https://discovery.url.edu/permalink/34CSUC_URL/1im36ta/alma991009794932906719
Tabla de Contenidos:
  • Front Cover
  • Computational Intelligence Methods for Sentiment Analysis in Natural Language Processing Applications
  • Copyright Page
  • Contents
  • List of contributors
  • Preface
  • 1 Role of machine learning in sentiment analysis: trends, challenges, and future directions
  • 1.1 Introduction
  • 1.1.1 Fundamentals of sentiment analysis
  • 1.1.2 Level of analysis
  • 1.1.2.1 Document level
  • 1.1.2.2 Sentence level
  • 1.1.2.3 Aspect-level analysis
  • 1.2 Related background
  • 1.2.1 Lexicon-based approach
  • 1.2.1.1 Dictionary-based approach
  • 1.2.1.2 Corpus-based approach
  • 1.2.2 Machine learning
  • 1.2.2.1 Unsupervised
  • 1.2.2.2 Supervised
  • 1.2.2.3 Semi-supervised
  • 1.2.3 Deep learning
  • 1.2.4 Hybrid approach
  • 1.3 Performance metrics
  • 1.4 Tools for sentiment analysis
  • 1.5 Trends of sentiment analysis
  • 1.5.1 Social media monitoring
  • 1.5.2 E-commerce
  • 1.5.3 Voice of customer
  • 1.5.4 Health care
  • 1.5.5 Financial services
  • 1.5.6 Political elections
  • 1.5.7 Recommendation system
  • 1.5.8 Employee retention
  • 1.6 Challenges
  • 1.6.1 Data gathering
  • 1.6.2 Sarcasm text
  • 1.6.3 Emojis
  • 1.6.4 Polarity
  • 1.6.5 Incomplete information
  • 1.6.6 Multimodal sentiment analysis
  • 1.6.7 Multilingual sentiment analysis
  • 1.7 Conclusion
  • 1.8 Future direction
  • References
  • 2 A comparative analysis of machine learning and deep learning techniques for aspect-based sentiment analysis
  • 2.1 Introduction
  • 2.2 Steps in sentiment analysis
  • 2.2.1 Data collection
  • 2.2.2 Data preprocessing
  • 2.2.3 Sentiment detection
  • 2.2.4 Sentiment classification
  • 2.2.5 Output preparation
  • 2.3 Applications of sentiment analysis
  • 2.3.1 Social media
  • 2.3.2 Business
  • 2.3.3 Pharmaceutical industry
  • 2.3.4 Politics
  • 2.4 Types of sentiment analysis
  • 2.5 Aspect-based sentiment analysis
  • 2.5.1 Machine learning-based approaches.
  • 2.5.1.1 Simple machine learning-based aspect-based sentiment analysis
  • 2.5.1.2 Machine learning approaches using hybrid technique and support vector machine
  • 2.5.1.3 Machine learning approaches using naïve Bayes methods
  • 2.5.2 Deep learning-based approaches
  • 2.5.2.1 Convolutional neural network-based aspect-based sentiment analysis
  • 2.5.2.2 Attention-based aspect-based sentiment analysis
  • 2.5.2.3 Long short-term memory -based aspect-based sentiment analysis
  • 2.6 Performance metrics
  • 2.7 Datasets
  • 2.7.1 Domain-specific dataset
  • 2.7.2 Language-specific datasets
  • 2.7.3 SemEval datasets
  • 2.8 Future research challenges
  • 2.8.1 Domain adaptation
  • 2.8.2 Multilingual application
  • 2.8.3 Unified model for multiple tasks
  • 2.8.4 Syntax-aware and position-aware model
  • 2.9 Conclusion
  • References
  • 3 A systematic survey on text-based dimensional sentiment analysis: advancements, challenges, and future directions
  • 3.1 Introduction
  • 3.1.1 Basic types of sentiment analysis tasks
  • 3.1.2 Need for multidimensional sentiment analysis
  • 3.2 Literature survey
  • 3.2.1 Types of sentiment analysis
  • 3.2.1.1 Multi-aspect sentiment analysis
  • 3.2.1.2 Multiclass sentiment analysis
  • 3.2.1.3 Multidimensional sentiment analysis
  • 3.2.1.4 Multidomain sentiment analysis
  • 3.2.1.5 Multi-lingual sentiment analysis
  • 3.2.1.6 Multi-modal sentiment analysis
  • 3.2.1.7 Multiscale sentiment analysis
  • 3.2.1.8 Multi-source sentiment analysis
  • 3.2.2 Framing survey queries for systematic literature review
  • 3.2.3 Advancements of dimensional sentiment analysis
  • 3.3 Observations drawn from the literature survey
  • 3.3.1 Responses to survey queries
  • 3.3.2 Levels of sentiment analysis
  • 3.3.2.1 Aspect-level sentiment analysis
  • 3.3.2.2 Phrase-level sentiment analysis
  • 3.3.2.3 Sentence-level sentiment analysis.
  • 3.3.2.4 Document-level sentiment analysis
  • 3.3.3 Steps to perform dimensional sentiment analysis
  • 3.4 Open issues and challenges in dimensional sentiment analysis
  • 3.5 Future directions
  • 3.6 Conclusion
  • References
  • 4 A model of time in natural linguistic reasoning
  • 4.1 Introduction
  • 4.2 Human biology of time
  • 4.2.1 Working hypotheses for time management
  • 4.2.2 Criteria for validity
  • 4.2.3 Materials and methods
  • 4.2.4 Biological foundations of time in human mind
  • 4.2.4.1 Animal encephalon structures involved
  • 4.2.4.2 Nuclei functioning as a storage of information
  • 4.2.4.3 Nuclei functioning as a time device
  • 4.2.4.4 Time perception and its alterations
  • 4.3 Evidence of timelines in the brain: time in linguistic reasoning
  • 4.4 Some clues and tests
  • 4.4.1 A device to model timelines
  • 4.4.2 A use case to test timelines
  • 4.4.3 Modeling the use case with NNR4
  • 4.4.3.1 First rule: timelessness out of C
  • 4.4.3.2 Second rule: cycles of use case timelines
  • 4.4.3.3 Third rule: poles in c according to timelines
  • 4.4.3.4 Fourth rule: the physical expression of timelines
  • 4.5 Conclusions and future work
  • References
  • 5 Hate speech detection using LSTM and explanation by LIME (local interpretable model-agnostic explanations)
  • 5.1 Introduction
  • 5.2 Bag of words
  • 5.2.1 Limitations of the Bag of Words
  • 5.2.2 Advantages of the Bag of Words
  • 5.3 Term frequency-inverse document frequency
  • 5.3.1 Why is TF-IDF?
  • 5.3.2 How is it calculated?
  • 5.3.3 Advantages and disadvantages of TF-IDF
  • 5.4 Glove-word embedding
  • 5.4.1 What is glove?
  • 5.5 Long short-term memory
  • 5.5.1 Structure of LSTM
  • 5.6 LIME-local interpretable model-agnostic explanations
  • 5.6.1 Working of LIME
  • 5.6.2 Explaining LSTM model predictions
  • 5.7 Code
  • References.
  • 6 Enhanced performance of drug review classification from social networks by improved ADASYN training and Natural Language ...
  • 6.1 Introduction
  • 6.2 Related works
  • 6.3 Proposed model
  • 6.3.1 Tokenization
  • 6.3.2 Stop words removal
  • 6.3.3 Normalization
  • 6.3.4 Feature extraction
  • 6.3.4.1 CountVectorizer
  • 6.3.4.2 Term frequency and inverse document frequency
  • 6.3.4.3 Word2vec
  • 6.3.4.4 Improved adaptive synthetic sampling (ADASYN)
  • 6.4 Results and discussion
  • 6.5 Conclusion
  • References
  • 7 Emotion detection from text data using machine learning for human behavior analysis
  • 7.1 Introduction
  • 7.1.1 Human behavior analysis
  • 7.1.1.1 Theories of emotion
  • 7.1.1.1.1 Social cognitive theory
  • 7.1.1.2 Theories of personality
  • 7.1.1.2.1 Social identity theory
  • 7.1.1.2.2 Self-determination theory
  • 7.1.2 Models of emotion
  • 7.1.3 Affective computing for emotion detection
  • 7.1.4 Natural language processing for emotion detection
  • 7.1.4.1 Applications
  • 7.1.4.2 Challenges
  • 7.2 Available tools and resources
  • 7.2.1 Tools for emotion detection
  • 7.2.2 Datasets
  • 7.2.3 Feature extraction tools
  • 7.3 Methods and materials
  • 7.3.1 Rule-based approaches
  • 7.3.1.1 Lexicon-based approach
  • 7.3.1.2 Keyword-based approach
  • 7.3.1.3 Regular expressions
  • 7.3.2 Statistical learning
  • 7.3.2.1 Machine learning-based approaches
  • 7.3.2.2 Deep learning methods
  • 7.3.2.3 Contextual adaptation
  • 7.3.2.4 Personalized bots
  • 7.3.3 Explainable AI for emotion detection
  • 7.4 Outlook
  • 7.4.1 Ethical considerations
  • 7.4.2 Limitations of NLP in emotion detection
  • 7.5 Conclusion
  • References
  • 8 Optimization of effectual sentiment analysis in film reviews using machine learning techniques
  • 8.1 Introduction
  • 8.2 Literature Survey
  • 8.3 Proposed System
  • 8.3.1 Hybrid algorithm Design
  • 8.3.2 Solution Representation.
  • 8.3.3 Improved Flowchart for C5.0BN
  • 8.3.4 Importance of Splitting Criteria in Decision Tree Algorithms
  • 8.4 Computational Experiments and Result Analysis
  • 8.4.1 Implementation of C5.0BN using scikit-learn
  • 8.4.2 Perform Test and Train Split
  • 8.4.3 Fit on Train Set
  • 8.4.4 Predict on Test Set
  • 8.4.5 Measure Accuracy of the Classifier
  • 8.4.5.1 Accuracy
  • 8.4.5.2 Memory used
  • 8.4.5.3 Training time
  • 8.4.5.4 Search time
  • 8.4.5.5 Error rate
  • 8.5 Conclusion
  • References
  • 9 Deep learning for double-negative detection in text data for customer feedback analysis on a product
  • 9.1 Introduction
  • 9.2 Related work
  • 9.3 Proposed methodology
  • 9.4 Experimental results and discussion
  • 9.5 Conclusion
  • References
  • 10 Sarcasm detection using deep learning in natural language processing
  • 10.1 Introduction
  • 10.1.1 What is sarcasm?
  • 10.1.2 Sentiment analysis
  • 10.1.3 Sarcasm detection
  • 10.2 Datasets
  • 10.2.1 News headline dataset for sarcasm detection
  • 10.2.2 Sarcasm corpus V2
  • 10.2.3 Sarcasm detection from Kaggle
  • 10.2.4 iSarcasm dataset
  • 10.3 Overall process of sarcasm detection
  • 10.4 Sarcasm detection and classification
  • 10.4.1 Deep neural network
  • 10.4.2 Convolutional neural networks
  • 10.4.3 BERT
  • 10.4.4 Long short-term memory
  • 10.4.5 Recurrent neural network
  • 10.4.6 BiLSTM
  • 10.5 Sarcasm detection: python code implementation
  • 10.6 Evaluation
  • 10.7 Results and discussion
  • 10.8 Conclusion
  • References
  • Further reading
  • 11 Abusive comment detection in Tamil using deep learning
  • 11.1 Introduction
  • 11.2 Related work
  • 11.2.1 Abusive comment detection in Tamil
  • 11.2.2 Hate speech and offensive content identification
  • 11.2.3 Sentiment analysis for Dravidian languages in code-mixed text
  • 11.2.4 Homophobia and transphobia detection in social media comments.
  • 11.3 Dataset description.

Ejemplares similares