Keywords

1 Introduction

Today, once a political corruption case takes place, it is rapidly viralized along the Internet where people can react by posting their opinions through widely used social networks such as Facebook and Twitter. Such audience reaction is clearly interesting and useful information for several institutions such as the government and of course for the opposite side. However, this information is complex to analyze as the audience employs stereotypes, metaphors, and ironies expressed in an informal language full of chat abbreviations, emoticons, and slang words hard to interpret.

In this paper, we present how content analysis can help us to uncover the hidden meaning of a message. Content analysis is a research technique for the objective, systematic and quantitative description of the manifest content of a given communication. We focus here on the automated analysis of two political corruption cases and its corresponding opinions through Twitter. The analysis period begins in September 2014 and ends in May 2015. The first one is the Penta case, concerned with the irregular financing for political campaigns, where a party from the right side was primarily involved. The second one is the Caval case, which is about influence peddling exerted by Natalia Compagnon and her husband Sebastián Dávalos, son of the president of the Chilean Republic, Michelle Bachelet. We present interesting results where the use of Content analysis allows us to easily process the social network information in order to provide clear feedback.

This paper is organized as follows: Sect. 2 presents the problem, results and discussion, followed by the conclusions and some lines of future directions.

2 Results and Discussion

Research in social networks began in the 90 s and one of the pioneers in the study of this research area were Wasserman and Faust [8]. This work can be seen as a framework for analyzing related data in social networks. However, limited work can be encountered in this context. Some related examples can be seen in [4] and in [1]. In the first one authors analyze social networks in the periods of the political campaign of Barack Obama, while the second one is focused on the wrong use of microblogging by mayors.

This article proposes the use of content analysis as a research technique to observe and reveal how the audience reacts to the publication of the news of corruption in Chile. The goal of Content Analysis is to fundamentally develop and process relevant data about conditions in which the text has been produced. According to Piñuel [7], content analysis aims to achieve the emergence of the latent meaning of the message, de-hiding the unapparent elements of every message. This research tool will allow us to make inferences from data collected and provide new knowledge of the visibilized reactions of a given audience.

Content analysis is considered as a privileged platform to access messages, since through a systematic and objective procedure we can describe the content of the messages [2]. The study of tweets offers us the opportunity to investigate the nature of the message, since it is possible to know how is the emitter and its associated thinking and ideologies [6]. As mentioned by Bartolomé [3] in this type of analysis we should avoid to make the following major mistake: removing the words of context. One word out of context says nothing about it nor about who produced it. Therefore, all our analysis will be carried out always taking into consideration who produced and communicated the message.

When content analysis is applied to messages, we are getting knowledge of who issues the message, as the discursive work is the product of his thinkings. Thus, judgments and comments issued by the producer allow us to decode their ideological structure, its context and its cultural heritage. After applying the content analysis to the tweets, as Piñuel [7] points out, we achieve the emergence of the latent sense of all social practice.

When analyzing the people reactions in the two corruption cases analyzed, we found the same act is repeated: The strategy used was the positive self-presentation and negative presentation of the other ones [5]. The reactions are performed by people who defend their political sector, which led to highlight corruption cases from the opposite side in order to diminish or soften the crime of their political sector.

In these cases we found that discussions through social networks are carried out just to remember the crimes committed by the opposite political side, without commenting on the current news published. Therefore, the strategy used relies in defending the own political sector, attacking the contrary. Thus the discussion is built around the axis Good\(\backslash \)Bad and most of reactions analyzed are stated in that line.

Another theme present in the reactions is the impunity of political crime. It is observed that there is a discontent among citizens about how politicians are treated w.r.t a crime in contrast to the rest of citizens. The existence of two citizen classes is manifested, a first one fully of privileges and impunity which most politicians belong to, and second class fully of complications with no privileges for most common citizens.

3 Conclusion

In this paper we have analyzed the reactions that people post on social networks about two major corruption cases in Chile. We found that regardless of political party, people tend to have the same reaction when defending their sector. In addition, reactions are unrelated to the news and the space is only used to criticize and besmirch the opposing side. No real discussion is presented, but only constant attacks from both sectors can be visualized. As future work we aim at analyzing other similar thematics to see if this same phenomenon is repeated. To study the form in which content analysis techniques can be integrated in text mining software could be another interesting direction for future work as well.