Conference PaperPDF Available

An approach to detect and analyze the impact of biased information sources in the social media

Authors:

Abstract

Abstract: The paper presumes that social media is an environment where local and small events may escalate into bigger and even global ones in a very short period of time. This is because social media offers opportunities for discussion of shared interest in way which cannot be controlled: everything that can be exposed will be exposed – for all intents and purposes. This possibility has also changed the landscape of discussions of controversial issue, such as foreign and security policy. Compared to traditional mass media, social media enable disclosing opinions without censorship. Nowadays people have access to online discussions, blogs and even websites entirely devoted to sharing negative information. It has been seen that, during crisis situations social media has become a major way of affecting people’s opinions. Consequently we are witnessing the rise of trolls – individual who shares inflammatory, extraneous or off-topic messages in social media, with the primary intent of provoking readers into an emotional response or of otherwise disrupting normal on-topic discussion. Based on the lack of censorship, on the one hand, and trolling behaviour, on the other, the paper aims to understand the rise and diffusion of extreme opinions in Twitter. This is a case study paper, where the analysed case is Twitter messages on Ukrainian crisis during 2014 written in Finnish language. The aim is to utilize sentiment analysis for the automatic detection of trolling behaviour. Sentiment analysis provides tools for strategic communications for the automatic analysis of social media discussions and to recognize opportunities for participating in the discussion at the most effective stage.
... This paper continues the work done by Paavola and Jalonen (2015), who examined whether sentiment analysis could be utilised in detecting trolling behaviour. Sentiment analysis refers to the automatic classification of messages into positive, negative, or neutral within a discussion topic. ...
... To facilitate analysis, a sentiment analysis tool (Paavola & Jalonen 2015) was further developed to detect message automation, which creates 'noise' in social media and makes it difficult to observe behavioural changes among human users of the social media. Paavola and Jalonen's work followed studies performed by Chu et al. (2012), Dickerson, Kagan, and Subrhamanian (2014), and Clark et al. (2015) in which bot detection systems were devised. ...
... Goals of the experimental case study were 1) to analyse how to detect fakeholders, and 2) to develop the sentiment analysis tool (Paavola & Jalonen 2015) to detect message automation. The essential issue of this study is to determine the properties indicating that a message is sent by a bot or by a cyborg. ...
Article
Full-text available
Social media has become a place for discussion and debate on controversial topics and, thus, provides an opportunity to influence public opinion. This possibility has given rise to a specific behaviour known as trolling, which can be found in almost every discussion that includes emotionally appealing topics. Trolling is a useful tool for any organisation willing to force a discussion off-track when one has no proper facts to back one’s arguments. Previous research has indicated that social media analytics tools can be utilised for automated detection of trolling. This paper provides tools for detecting message automation utilized in trolling.
... The annotated data set was not published. Paavola and Jalonen (2015) used sentiment analysis in order to detect trolling behavior in tweets in Finnish during the 2014 Ukrainian crisis. They used a social media analysis tool developed in the NEMO project detecting the polarity (positive-neutral-negative) of the messages. ...
... Jussila et al. (2017) investigated the reliability of two sentiment analysis tools for Finnish when compared with human evaluators. The two analysis tools were the SentiStrength (Thelwall et al., 2010(Thelwall et al., , 2012 and the Nemo Sentiment and Data Analyzer (Paavola and Jalonen, 2015). The Nemo Sentiment and Data Analyzer tool can also be used to collect tweets and it was used to collect a set of 509 tweets in Finnish. ...
Preprint
Full-text available
Sentiment analysis and opinion mining is an important task with obvious application areas in social media, e.g. when indicating hate speech and fake news. In our survey of previous work, we note that there is no large-scale social media data set with sentiment polarity annotations for Finnish. This publications aims to remedy this shortcoming by introducing a 27,000 sentence data set annotated independently with sentiment polarity by three native annotators. We had the same three annotators for the whole data set, which provides a unique opportunity for further studies of annotator behaviour over time. We analyse their inter-annotator agreement and provide two baselines to validate the usefulness of the data set.
... More broadly, Kriplean et al. argue that systems should be designed to reward participants not only to contribute but also to listen well (e.g., restating other people's comments [9], providing pros and cons [8]). One of the challenges of conducting a debate online, is the fact that they can lead to trolling behaviors, where people with extreme opinions take over the conversation [13]. ...
Conference Paper
Full-text available
Students increasingly have access to information that can be posted by anyone without being vetted, and it becomes vital to support students in evaluating claims through debate and critical thinking. To address this issue, this paper designs and evaluates a light-weight but effective protocol for supporting debate in a classroom activity with university students. It evaluated participants' beliefs on controversial topics (e.g., homeopathy) before and after briefly learning about critical thinking tools, posting arguments, and critically evaluating the arguments of peers. The findings suggest that this intervention led to a statistically significant belief change, and that this change was in the direction of the position best supported by evidence. Consequently, this work in progress presents a constructive approach to scaffold debates in the classroom and beyond.
... Rumour detection. Recently, more computing studies have investigated the emergence of rumours, but they stay at the level of a specific rumour, as in Fig 1 [32][33][34][35][36][37][38][39]. Contrary to these studies, our goal is to analyse any kind of rumour and a corpus of rumours. ...
Article
Full-text available
Rumour is an old social phenomenon used in politics and other public spaces. It has been studied for only hundred years by sociologists and psychologists by qualitative means. Social media platforms open new opportunities to improve quantitative analyses. We scanned all scientific literature to find relevant features. We made a quantitative screening of some specific rumours (in French and in English). Firstly, we identified some sources of information to find them. Secondly, we compiled different reference, rumouring and event datasets. Thirdly, we considered two facets of a rumour: the way it can spread to other users, and the syntagmatic content that may or may not be specific for a rumour. We found 53 features, clustered into six categories, which are able to describe a rumour message. The spread of a rumour is multi-harmonic having different frequencies and spikes, and can survive several years. Combinations of words (n-grams and skip-grams) are not typical of expressivity between rumours and news but study of lexical transition from a time period to the next goes in the sense of transmission pattern as described by Allport theory of transmission. A rumour can be interpreted as a speech act but with transmission patterns.
Chapter
The potential of social media in marketing communications is widely acknowledged, but in terms of making use of social media enabled analytics there are still uncovered possibilities for marketers. The focus of this paper is on analyzing the customers' affective experiences appearing in the social media content. Through action design research, a framework that enables to analyze affective experiences from social media content is developed. The novelty of the framework is that it takes into account the different emotion families as well as the intensity of the affective experience, taking a one step further of the generally used sentiment analysis techniques in social media context. Through this kind of framework, the marketers are able to better catch even the weak emotional signals of the customers and to guide the customer to more valuable emotional path. This paper presents the developed framework and its pilot testing, carried out as part of a wider research process involving two research projects and researchers from three different universities.
Chapter
This chapter discusses the transition of the Baltic Sea Region (BSR) from a hinge between Russia and the West to a rhizome—from a fixed and linear point of connection between two geopolitical areas to a polydimensional network of nodal interactivity. To this end, the chapter takes as its primary focal points the rhizomatic1 information channels and mechanisms that currently exist for both communication and conflict in the region, as well as the way in which the East–West dialogue can be fostered in the Baltic context, and the importance of rhizomatic technologies for future interactions in the BSR. In tracing this development, the chapter posits that (1) an ideological misunderstanding has emerged between the West and Russia, inflamed by the Ukraine war and exacerbated by the wave of refugees into Europe; (2) the media, especially social media in the information domain, play a significant role in compounding this misunderstanding, but also play a key role in keeping the information channels open, especially in the security domain; (3) the BSR countries may still have a role in resolving the misunderstanding thanks to their traditional hinge-like connection to Russia, ranging from geopolitics to information politics; and (4) the misunderstanding that exists not only poses a risk to the rhizomatic information networks but may also point the way to a new kind of mobile virtual neighbouring. Indeed, rhizomatic information hybridity and mobile virtual perspectives offer the possibility for more diffused and heterogeneous understandings than those that are confined within state boundaries (Browning and Joenniemi 2007, 17).
Conference Paper
Full-text available
Social media offers rich data sources for companies that want to understand how they are perceived by their stakeholders. Sentiment analysis over Twitter can produce information about people’s feelings toward their brand, business, and directors (Saif et al. 2012). Based on this information, companies can take actions to enhance their customer experiences and perceived brand value. This study investigates the reliability and perceived value of two sentiment analysis tools developed to understand Finnish language, in contrast to human evaluators. For this purpose, a dataset of tweets from a Finnish software company was collected. For evaluating reliability Krippendorff’s α (Krippendorff 2007) is computed. Perceived value of the automatic and human evaluator classified sentiment is evaluated by interviewing the case company representatives. The results point out that the analysis carried out by the human evaluators was perceived more valuable by the company representatives than the automatic analysis, due to different granulation level of the analysis. Compared to the automatic analysis, the human evaluators were able to put the identified emotions from the tweets better into a context, which in turn diminished the potential misinterpretation of who was the target of the most negative tweets.
Poster
Full-text available
The purpose of the study is to present a framework for a deeper analysis of affecFve experiences from social media content.
Article
Full-text available
Article
Twitter is a popular microblogging service that is used to read and write millions of short messages on any topic within a 140-character limit. Popular or influential users tweet their status and are retweeted, mentioned, or replied to by their audience. Sentiment analysis of the tweets by popular users and their audience reveals whether the audience is favorable to popular users. We analyzed over 3,000,000 tweets mentioning or replying to the 13 most influential users to determine audience sentiment. Twitter messages reflect the landscape of sentiment toward its most popular users. We used the sentiment analysis technique as a valid popularity indicator or measure. First, we distinguished between the positive and negative audiences of popular users. Second, we found that the sentiments expressed in the tweets by popular users influenced the sentiment of their audience. Third, from the above two findings we developed a positive-negative measure for this influence. Finally, using a Granger causality analysis, we found that the time-series-based positive-negative sentiment change of the audience was related to the real-world sentiment landscape of popular users. We believe that the positive-negative influence measure between popular users and their audience provides new insights into the influence of a user and is related to the real world.
Article
The news feeds of two U.S. politicians' Facebook sites were examined across 22 months leading up to an election 6 in order to explore changes in social-network-mediated public political discourse over time. Changes over time were observed 7 in who was being addressed and in the affective valence of comments. A complex flow of attention between in-group and 8 out-group concerns was observed with in-group comments dominant both in early and late phases. Also, positive comments 9 decreased and negative comments increased over time. These phenomena, dubbed "reflection-to-selection" and "converging 10 sentiment", were refined to explain the observed nonlinearities. The flow of rational versus affective comments in politicians' 11 Facebook data across time was also explored. Comments reflecting cognition were more prevalent at all times than comments 12 reflecting affect, but their distribution also varied in complex ways over time. Finally, the concept of "potential public sphere" 13 in contrast to "realized public sphere" in virtual spaces is introduced.
Article
It is clear that customer-to-customer contact through informal social networking and more formal company-sponsored platforms, such as discussion forums, is an increasingly integral element in building brand communities. There are many benefits to this interaction, such as nurturing brand champions and ‘super users,’ and reduced service costs through customer-to-customer solutions for product problems. However, there are also hazards inherent in these largely unregulated communities, such as the potential damage of widely spread negative information, which may be based on fact or on malicious intent. Herein, we summarize the results of several years of research examining these communities in an attempt to understand why they succeed, what benefits can be extracted from them, and—in particular—how negative information emerging in these environments can be strategically managed. Based on a series of quantitative and qualitative studies, we identified several key drivers of online brand community success (i.e., intervention, conversion, value creation, and harvesting) and the different combinations of community players who must collaborate to achieve such success. Delving more deeply into the issue of negative information management, we find that the topic being discussed (i.e., core versus augmented product) and the validity of the claim greatly influence a firm's appropriate strategic response. Throughout this article, we offer managerial guidance on the most effective ways to develop primary brand communities that encourage loyalty, purchases, and positive word of mouth.
Article
We present an analysis of the properties of a social network, formed by users participating in one of the most popular Polish political discussion forums, during a period of two years. We find that despite communication motivated mostly by controversy and negative emotions, the resulting social network is remarkably similar to networks found in discussion boards based on similarity of views. All collected messages have been analyzed to determine their content type, author's political sympathies and emotion level. This has allowed us to discover almost complete stability of the political sympathies of the users writing comments during the studied period. Using the data on reader evaluation of the comments we present estimates of the political sympathies of the "silent majority" of forum readers and found that they follow closely the distribution of sympathies of comment authors. The data from the observations are used as the input for an agent based computer simulation model, reproducing many observed statistical characteristics of the forum.
Article
When the change process accelerates, the meaning of the logic of cause and effect in living through this blurring reality clearly diminishes. We have to tell the story of our reality and future both to ourselves and to society with new concepts, words and metaphors. While the importance of instant experience and transience grows, our in-depth understanding of the logic of time gradually crumbles away. The fervency of action widens the limits of reality in a fundamental way, even though our personal human capacity to receive and handle information and experience emotions and events has not changed in the one million years of human history.
Article
Abstract Despite evidence that anger is routinely expressed over the Internet via weblogs, social networking Web sites, and other venues, no published research has explored the way in which anger is experienced and expressed online. Consequently, we know very little about how anger is experienced in such settings. Two studies were conducted to explore how people experience and express their anger on a particular type of Web site, known as a rant-site. Study 1 surveyed rant-site visitors to better understand the perceived value of the Web sites and found that while they become relaxed immediately after posting, they also experience more anger than most and express their anger in maladaptive ways. Study 2 explored the emotional impact of reading and writing rants and found that for most participants, reading and writing rants were associated with negative shifts in mood.
Article
The present study examines the influence of social context on the use of emoticons in Internet communication. Secondary school students (N = 158) responded to short internet chats. Social context (task-oriented vs. socio-emotional) and valence of the context (positive vs. negative) were manipulated in these chats. Participants were permitted to respond with text, emoticon or a combination of both. Results showed that participants used more emoticons in socio-emotional than in task-oriented social contexts. Furthermore, students used more positive emoticons in positive contexts and more negative emoticons in negative contexts. An interaction was found between valence and kind of context; in negative, task-oriented contexts subjects used the least emoticons. Results are related to research about the expression of emotions in face-to-face interaction.