Main description:

In this book common sense computing techniques are further developed and applied to bridge the semantic gap between word-level natural language data and the concept-level opinions conveyed by these. In particular, the ensemble application of graph mining and multi-dimensionality reduction techniques is exploited on two common sense knowledge bases to develop a novel intelligent engine for open-domain opinion mining and sentiment analysis. The proposed approach, termed sentic computing, performs a clause-level semantic analysis of text, which allows the inference of both the conceptual and emotional information associated with natural language opinions and, hence, a more efficient passage from (unstructured) textual information to (structured) machine-processable data.

Feature:

Represents the first comprehensive review of Sentic Computing, state-of-the-art approach to opinion mining and sentiment analysis (see http://en.wikipedia.org/wiki/Sentiment_analysis)

A special chapter on cognitive and affective modeling for natural language understanding

Includes tips on different strategies (techniques, online resources, datasets, etc.) to opinion mining and sentiment analysis

Back cover:

In this book common sense computing techniques are further developed and applied to bridge the semantic gap between word-level natural language data and the concept-level opinions conveyed by these. In particular, the ensemble application of graph mining and multi-dimensionality reduction techniques is exploited on two common sense knowledge bases to develop a novel intelligent engine for open-domain opinion mining and sentiment analysis. The proposed approach, termed sentic computing, performs a clause-level semantic analysis of text, which allows the inference of both the conceptual and emotional information associated with natural language opinions and, hence, a more efficient passage from (unstructured) textual information to (structured) machine-processable data.

Contents:

Introduction

1.1 Sentic Computing
  1.1.1 Motivations
  1.1.2 Aims
  1.1.3 Methodology

Background

2.1 Opinion Mining and Sentiment Analysis
  2.1.1 The Buzz Mechanism
  2.1.2 Origins and Peculiarities
  2.1.3 Sub-Tasks
2.2 Main Approaches to Opinion Mining
  2.2.1 From Heuristics to Discourse Structure
  2.2.2 From Coarse to Fine Grained
  2.2.3 From Keywords to Concepts
2.3 Towards Machines with Common Sense
  2.3.1 The Importance of Common Sense
  2.3.2 Knowledge Representation
  2.3.3 From Logical Inference to Digital Intuition
2.4 Conclusions

Techniques

3.1 Affective Blending: Enabling Emotion-Sensitive Inference
  3.1.1 AffectNet
  3.1.2 AffectiveSpace
3.2 Affective Categorisation: Modelling Human Emotions
  3.2.1 Categorical Versus Dimensional Approaches
  3.2.2 The Hourglass of Emotions
3.3 Sentic Medoids: Clustering Affective Common Sense Concepts
  3.3.1 Partitioning Around Medoids
  3.3.2 Centroid Selection
3.4 Sentic Activation: A Two-Level Affective Reasoning Framework
  3.4.1 Unconscious Reasoning
  3.4.2 Conscious Reasoning
3.5 Sentic Panalogy: Switching Between Different Ways to Think
  3.5.1 Changing Reasoning Strategies
  3.5.2 Changing Reasoning Foci
3.6 Conclusions

Tools

4.1 SenticNet: A Semantic Resource for Opinion Mining
  4.1.1 Building SenticNet
  4.1.2 Working with SenticNet
4.2 Sentic Neural Networks: Brain-Inspired Affective Reasoning
  4.2.1 Discrete Versus Continuous Approach
  4.2.2 Affective Learning
4.3 Open Mind Common Sentics: An Emotion-Sensitive IUI
  4.3.1 Games for Knowledge Acquisition
  4.3.2 Collecting Affective Common Sense Knowledge
4.4 Isanette: A Common and Common Sense Knowledge Base
  4.4.1 Probase
  4.4.2 Building the Instance-Concept Matrix
4.5 Opinion Mining Engine: Structuring the Unstructured
  4.5.1 Constitutive Modules
  4.5.2 Evaluation
4.6 Conclusions

Applications

5.1 Development of Social Web Systems
  5.1.1 Troll Filtering
  5.1.2 Social Media Marketing
  5.1.3 Sentic Album
5.2 Development of HCI Systems
  5.2.1 Sentic Avatar
  5.2.2 Sentic Chat
  5.2.3 Sentic Corner
5.3 Development of E-Health Systems
  5.3.1 Crowd Validation
  5.3.2 Sentic PROMs
5.4 Conclusions

Concluding Remarks

6.1 Summary of Contributions
  6.1.1 Techniques
  6.1.2 Tools
  6.1.3 Applications
6.2 Limitations and Future Work
  6.2.1 Limitations
  6.2.2 Future Work
6.3 Conclusions

References