The great diffusion of social networks and their role in modern society are among the more interesting novelties in recent years, capturing the interest of researchers, journalists, companies, and governments. The dense interconnection that often arises among active users generates a discussion space that is able to motivate and involve individuals, linking people with common objectives and facilitating diverse forms of collective action. This gives rise to what is called “individualism on the net”: instead of always counting on a single reference community, thanks to the social networks it becomes possible to be stimulated by moving among more people and resources, which are often heterogeneous. Social networks are therefore creating a digital revolution. The most interesting aspect of this change is not solely related to the possibility of promoting political participation and activism. This social revolution influences the lives of every individual. It is the freedom to express ourselves, to have our own space in which we can be ourselves, or to be who we would like to be, with few limits and barriers.

The social network revolution enables us to talk about our emotions and opinions not only to ourselves but especially to those around us, interacting with them, opening a window on the worlds of others, and snooping into their lives. Paradoxically, all this happens while we are living in a society where it is even more difficult to know the names of our neighbors, and where the right to privacy becomes an imperative to which we submit. Because of social media, we effectively end up telling the whole (or most) of our life: happiness at the birth of a child, anger at a train delay, pre-Christmas shopping, or the choice made in the secrecy of the voting booth. It is then not surprising that researchers started to discuss the methods to capture this vast sea of information. The data on the net, if properly collected and analyzed, allows us not only to understand and explain many complex social phenomena but also to predict them. In this context, sentiment analysis tries to make evident what people think by providing representations, models, and algorithms able to move from “simple unstructured text” to “complex insight.” In this book we want to provide a comprehensive understanding of this topic, with a particular emphasis on its role in social networks.

Sentiment Analysis in Social Networks is first of all directed at researchers with computer science, mathematics, and statistics skills both in academia and in industry. Researchers in academia can benefit from a consolidated knowledge background to create innovative solutions tuned in the challenging environment of social networks. On the other hand, companies and organizations can take advantage of the most recent state of the art to innovate processes, products, and services and therefore increase their competitiveness.

The book is organized as follows: Chapter 1 is intended to make the reader aware of the challenges related to sentiment analysis in social networks and prepare the reader for a full and precise understanding of the problems and solutions presented in the next chapters.

Chapter 2 introduces the psychological and sociological processes underlying social network interactions. The chapter starts by highlighting the differences and specific features that characterize online social network dynamics and finally points out how this understanding can be effectively exploited by sentiment analysis methodological approaches.

Chapter 3 describes the role of semantics in sentiment analysis by analyzing two main perspectives: the way semantics is encoded in sentiment resources, such as lexica, corpora, and ontologies, and the way it is used by automatic systems that perform sentiment analysis on social network data.

Chapter 4 highlights the main issues related to the interoperability of language resources for sentiment analysis. After an introduction to the linked data perspective for semantic modeling of sentiment and emotions, the chapter presents a linked data model based on two vocabularies: Marl and Onyx.

Chapter 5 presents the most recent state of the art with regard to the affective characterization of sentiment in social networks, then highlighting how affective reasoning can be employed for the development of applications in the context of big social data analysis.

Chapter 6 discusses a machine learning approach to sentiment analysis in social networks, by distinguishing between supervised, unsupervised, and semisupervised models and highlighting the potential leverage that the network structure can have.

Chapter 7 introduces irony and sarcasm when dealing with informal text on the network, underlining the difficulties that these figures of speech can cause for understanding the real sentiment of everyday communications.

Chapter 8 presents a comprehensive overview of a novel task related to suggestion mining from opinionated text. Various aspects are discussed, including the analysis of suggestions appearing in reviews, the relation between sentiments and suggestions, and the most recent methods for addressing the problem.

Chapter 9 introduces the problem of opinion spam detection. After an overview of the problem, different techniques to leverage the relationships between different entities in the network are presented for the detection of both opinion spam and opinion spammers.

Chapter 10 presents a comprehensive analysis of the state of the art related to opinion leader detection strategies and the associated challenges. A critical discussion of the advantages, limitations, and suitability of the approaches is provided, illustrating the potential for generating added value in different scenarios.

Chapter 11 surveys the techniques for both extractive and abstractive summarization of opinion-filled text, including a discussion of summarization evaluation. Then approaches to show opinion summaries to the users in the form of visualizations are discussed, including several interactive solutions.

In the following chapters, contributions from small- and medium-sized enterprises related to their solutions to sentiment analysis in social networks are presented. A comprehensive review of software tools and environments (both proprietary and open source) is presented and some case studies are outlined.

Chapter 12 presents SpagoBI, which is an entirely open source business intelligence suite. It offers a wide range of analytical tools to perform reporting, multidimensional analysis, key performance indicator management, visualization through charts, dashboards and cockpits, ad hoc querying and reporting, location intelligence, and network analysis.

Chapter 13 details SOMA, which is a smart social customer relationship management tool for companies aimed at monitoring and dealing with consumers’ complaints and interactions in social networks, with a particular focus on both sentiment and emotions.

Chapter 14 describes KRC’s proprietary social and digital media analytics suite, which is able to combine human intelligence with advanced predictive models.

Chapter 15 presents TheySay, which is a multidimensional opinion streaming architecture, powered by a large-scale custom natural language processing pipeline for sentiment, affect, and irrealis analysis used to monitor, quantify, and estimate the impact of corporate announcements to provide rich price-sensitive feedback and insights.

Finally, Chapter 16 presents a final discussion, concluding with some thoughts and opinions about future directions.