To read the full-text of this research, you can request a copy directly from the authors.
Recent rises in political polarization across the globe are often ascribed to algorithmic content filtering on social media, news platforms, or search engines. The widespread usage of news recommendation systems (NRS) is theorized to drive users in homogenous information environments and, thereby, drive affective, ideological, and perceived polarization. To test this assumption, we conducted an online experiment ( n = 750) with running algorithms that enriches content-based NRS with negative or neutral sentiment. Our experiment finds only limited evidence for polarization effects of content-based NRS. Nevertheless, the time spent with an NRS and its recommended articles seems to play a crucial role as a moderator of polarization. The longer participants were using an NRS enriched with negative sentiment, the more they got affectively polarized, whereas participants using an NRS incorporating balanced sentiment ideologically depolarized over time. Implications for future research are discussed.
To read the full-text of this research, you can request a copy directly from the authors.
... In contrast, NRSs that recommend politically diversity-optimized news instead of accuracy-optimized news are related to a higher tolerance for opposing views . Likewise, it was shown that spending more time with a content-based NRS accentuating negative sentiment increases affective polarization and that an NRS with balanced sentiment is related to decreasing ideological polarization the longer participants used the NRS . ...
... An earlier version of the German dataset  has been used in  and  to examine sentiment and stance recommender bias, and to investigate polarization and filter bubble creation through online studies, respectively. While in this study we have analyzed political biases of recommenders only from the perspective of users' click histories, in the future their political information can be further explored in conjunction with the models' predictions to understand whether any existing biases are reinforced or amplified by the algorithms. ...
News recommendation plays a critical role in shaping the public's worldviews through the way in which it filters and disseminates information about different topics. Given the crucial impact that media plays in opinion formation, especially for sensitive topics, understanding the effects of personalized recommendation beyond accuracy has become essential in today's digital society. In this work, we present NeMig, a bilingual news collection on the topic of migration, and corresponding rich user data. In comparison to existing news recommendation datasets, which comprise a large variety of monolingual news, NeMig covers articles on a single controversial topic, published in both Germany and the US. We annotate the sentiment polarization of the articles and the political leanings of the media outlets, in addition to extracting subtopics and named entities disambiguated through Wikidata. These features can be used to analyze the effects of algorithmic news curation beyond accuracy-based performance, such as recommender biases and the creation of filter bubbles. We construct domain-specific knowledge graphs from the news text and metadata, thus encoding knowledge-level connections between articles. Importantly, while existing datasets include only click behavior, we collect user socio-demographic and political information in addition to explicit click feedback. We demonstrate the utility of NeMig through experiments on the tasks of news recommenders benchmarking, analysis of biases in recommenders, and news trends analysis. NeMig aims to provide a useful resource for the news recommendation community and to foster interdisciplinary research into the multidimensional effects of algorithmic news curation.
Faced with the ongoing evolution of the media landscape, media and communication science is constant-ly asking itself new questions concerning the tension between stability and change in social communication. In this context, many relevant topics exist for which there are still no publications that would systematically evaluate what we know (empirically and non-empirically) so far and what conclusions can be drawn from existing knowledge. This volume aims to provide systematic answers to important current or continually relevant questions in the field, with contributions that are entirely focused on a specific question, thus leaving room for thorough arguments. The volume is dedicated to the memory of Professor Wolfram Peiser. With contributions by Hans-Bernd Brosius, Felix Frey, Romy Fröhlich, Christina Holtz-Bacha, Benjamin Krämer, Philipp Müller, Christoph Neuberger, Carsten Reinemann, Anna-Luisa Sacher, Johanna Schindler, Klaus Schönbach and Cornelia Wallner.
In 2020, societies debated the use of government restrictions on public life to stem the COVID-19 pandemic. Many of these debates took place online. The Internet enables people to come into contact with like-minded content. Algorithms based on collaborative filtering can contribute to this process and might lead to homogenous like-minded online environments that contribute to a polarisation of society. This article therefore examines the effects of (1) like-minded versus opposing online environments, which were (2) randomly versus algorithmically curated. Data from a between-subject experiment embedded in a two-wave panel survey of German citizens (n = 318) show that attitude polarisation as well as affective polarisation are largely independent of exposure to different online environments. Moreover, the results indicate that polarised attitudes of supporters and opponents of the COVID-19-related restrictions relate to varying degrees of beliefs in the importance of silencing people with opposing opinions: While supporters’ polarised attitudes are positively related to the belief in the importance of silencing others, opponents’ polarised attitudes are rather negatively related to such beliefs.
Recommender systems have been applied in a wide range of domains such as e-commerce, media, banking, and utilities. This kind of system provides personalized suggestions based on large amounts of data to increase user satisfaction. These suggestions help client select products, while organizations can increase the consumption of a product. In the case of social data, sentiment analysis can help gain better understanding of a user’s attitudes, opinions and emotions, which is beneficial to integrate in recommender systems for achieving higher recommendation reliability. On the one hand, this information can be used to complement explicit ratings given to products by users. On the other hand, sentiment analysis of items that can be derived from online news services, blogs, social media or even from the recommender systems themselves is seen as capable of providing better recommendations to users. In this study, we present and evaluate a recommendation approach that integrates sentiment analysis into collaborative filtering methods. The recommender system proposal is based on an adaptive architecture, which includes improved techniques for feature extraction and deep learning models based on sentiment analysis. The results of the empirical study performed with two popular datasets show that sentiment–based deep learning models and collaborative filtering methods can significantly improve the recommender system’s performance.
Selective exposure to likeminded political viewpoints on algorithmic social media platforms is considered a potential source of polarization of public opinion. We still know little about the proposed mechanism or how potential reinforcement of specific attitudes affects citizens’ political behavior, especially in a nonelectoral context. Focusing on the issue of immigration during the refugee influx to Europe in autumn 2015, this study investigates the effects of social media usage on attitude reinforcement, connecting it to political participation in refugee-related activities. A panel study conducted among Danish citizens (n = 847) reveals that frequent social media usage reinforces existing attitudes and mobilizes political participation. However, citizens who become more extreme in their attitude toward immigration over time are found to be less likely to become politically active regarding this specific issue.
Does exposure to social media polarize users or simply sort out like-minded voters based on their preexisting beliefs? In this paper, we conduct three survey experiments to assess the direct and unconditioned effect of exposure to tweets on perceived ideological polarization of candidates and parties. We show that subjects treated with negative tweets see greater ideological distance between presidential nominees and between their parties. We also demonstrate that polarization increases with processing time. We demonstrate a social media effect on perceived polarization beyond that due to the self-selection of like-minded users into different media communities. We explain our results as the result of social media frames that increase contrast effects between voters and candidates.
The article contributes both conceptually and methodologically to the study of online news consumption by introducing new approaches to measuring user information behaviour and proposing a typology of users based on their click behaviour. Using as a case study two online outlets of large national newspapers, it employs computational approaches to detect patterns in time- and content-based user interactions with news content based on clickstream data. The analysis of interactions detects several distinct timelines of news consumption and scrutinises how users switch between news topics during reading sessions. Using clustering analysis, the article then identifies several types of news readers (e.g. samplers, gourmets) and examines their news diets. The results point out the limited variation in topical composition of the news diets between different types of readers and the tendency of these diets to align with the news supply patterns (i.e. the average distribution of topics covered by the outlet).
This study examines algorithm effects on user opinion, utilizing a real-world recommender algorithm of a highly popular video-sharing platform, YouTube. We experimentally manipulate user search/watch history by our custom programming. A controlled laboratory experiment is then conducted to examine whether exposure to algorithmically recommended content reinforces and polarizes political opinions. Results suggest that political self-reinforcement, as indicated by the political emotion-ideology alignment, and affective polarization are heightened by political videos – selected by the YouTube recommender algorithm – based on participants’ own search preferences. Suggestions for how to reduce algorithm-induced political polarization and implications of algorithmic personalization for democracy are discussed.
What accounts for the prevalence of negative news content? One answer may lie in the tendency for humans to react more strongly to negative than positive information. “Negativity biases” in human cognition and behavior are well documented, but existing research is based on small Anglo-American samples and stimuli that are only tangentially related to our political world. This work accordingly reports results from a 17-country, 6-continent experimental study examining psychophysiological reactions to real video news content. Results offer the most comprehensive cross-national demonstration of negativity biases to date, but they also serve to highlight considerable individual-level variation in responsiveness to news content. Insofar as our results make clear the pervasiveness of negativity biases on average, they help account for the tendency for audience-seeking news around the world to be predominantly negative. Insofar as our results highlight individual-level variation, however, they highlight the potential for more positive content, and suggest that there may be reason to reconsider the conventional journalistic wisdom that “if it bleeds, it leads.”
Inwiefern digitale Medien politische Prozesse beeinflussen ist eine intensiv diskutierte Frage inner- und außerhalb der Politikwissenschaft. Besondere Prominenz in dieser Debatte hat dabei die Hypothese sogenannter „Echokammern“ gewonnen, wonach digitale Medien ihre Nutzer darin bestärken, insbesondere solche Nachrichten zu beziehen, deren politische Positionierung sie teilen, und dadurch zu einer gesellschaftlichen Polarisierung beitragen. Während Echokammern in der öffentlichen Debatte zumeist unkritisch als gegeben betrachtet werden, wird das Konzept im wissenschaftlichen Diskurs zunehmend hinterfragt. Als Herausforderungen erweisen sich dabei eine schwache theoretische Aufarbeitung des Phänomens, ein stark zersplittertes Forschungsfeld und eine mangelnde Generalisierbarkeit von Forschungsergebnissen aufgrund des primären Fokus auf den US-amerikanischen Kontext. Der vorliegende Beitrag begegnet diesen Problemstellungen und gibt einen detaillierten Überblick über das Forschungsfeld. Der Literaturüberblick trägt dabei zur theoretischen Erfassung des Untersuchungsgegenstands bei, insbesondere durch eine explizite Differenzierung zwischen Fragmentierung und Polarisierung, und berücksichtigt außerdem länderspezifische Variationen. Insgesamt kommt dieser Überblick zu dem Schluss, dass die im öffentlichen Diskurs geäußerte Furcht vor einer gesamtgesellschaftlichen Fragmentierung durch digitale Medien und einer damit verbundenen politischen Polarisierung empirisch nicht unterstützt wird. So ist aufbauend auf die bisherige Forschung keine Fragmentierung öffentlicher Aufmerksamkeit entlang politischer Präferenzen feststellbar. Auch auf der Wirkungsebene der Polarisierung sprechen die bisherigen Erkenntnisse gegen die vereinfachten Annahmen der Echokammer-Hypothese. Dennoch sind die bisherigen wissenschaftlichen Befunde aufgrund von Limitationen im Datenzugang noch nicht umfassend genug. Der Beitrag verdeutlicht, dass die politische Kommunikationsforschung insbesondere von innovativen, extern validen Designs und komparativer Forschung außerhalb des US-Kontexts profitieren würde.
This article proposes a definition of alternative news media and suggests routes for further research. It complements and extends previous conceptualizations in research on alternative media and outlines an umbrella definition of this phenomenon aimed to inspire contemporary research and scholarly debate. Previous research has been guided by a ‘progressive’ perspective as a form of resistance against ‘bourgeois’ hegemonic discourse. Such normative evaluations have in turn limited how the phenomenon has been studied empirically, by limiting the scope of research so that important contemporary phenomena fall outside the theoretical map. Conceptualizing alternative news media in the present hybrid and polarized media environment, we first propose a non-normative, multilevel relational definition: Alternative news media position themselves as correctives of the mainstream news media, as expressed in editorial agendas or statements and/or are perceived as such by their audiences or third-parties. This counter-hegemonic alternativeness can emerge on the macro level of societal function, the meso-level of organizations and/or the micro level of news content and producers. Second, demonstrating why this umbrella definition is fruitful in the changing media environment characterized by boundary struggles, crisis in legacy news media and mushrooming of alternative news outlets, we highlight research gaps and propose future research.
Social media sites are often blamed for exacerbating political polarization by creating “echo chambers” that prevent people from being exposed to information that contradicts their preexisting beliefs. We conducted a field experiment that offered a large group of Democrats and Republicans financial compensation to follow bots that retweeted messages by elected officials and opinion leaders with opposing political views. Republican participants expressed substantially more conservative views after following a liberal Twitter bot, whereas Democrats’ attitudes became slightly more liberal after following a conservative Twitter bot—although this effect was not statistically significant. Despite several limitations, this study has important implications for the emerging field of computational social science and ongoing efforts to reduce political polarization online.
Online algorithms have received much blame for polarizing emotions during the 2016 U.S. presidential election. We use transfer entropy to measure directed information flows from human emotions to YouTube’s video recommendation engine, and back, from recommended videos to users’ emotions. We find that algorithmic recommendations communicate a statistically significant amount of positive and negative affect to humans. Joy is prevalent in emotional polarization, while sadness and fear play significant roles in emotional convergence. These findings can help to design more socially responsible algorithms by starting to focus on the emotional content of algorithmic recommendations. Employing a computational-experimental mixed method approach, the study serves as a demonstration of how the mathematical theory of communication can be used both to quantify human-machine communication, and to test hypotheses in the social sciences.
Based on a quantitative content analysis of political actors’ Facebook posts (N = 1915), this study investigates profile-level and post-level drivers of user engagement (comments, likes, and shares) by employing a multilevel approach. For the first time in extant research, we also examine the factors that drive political actors to react to user comments. Findings indicate that the number of followers, the use of an official fan profile, and party
vote share were negatively related to political actors’ reactions to user comments. Furthermore, party profiles were least successful in stimulating user engagement. On the post level, we found that reasoning, post length, and references to competitive political actors have the potential to increase different types of user engagement. Negative, but not positive tonality increased user engagement and positive emotional expressions had a stronger effect on user engagement than negative emotions. Furthermore, humorous posts were more likely to be commented, liked, or shared, while mobilization cues had predominantly negative effects on user engagement.
In the debate about filter bubbles caused by algorithmic news recommendation, the conceptualization of the two core concepts in this debate, diversity and algorithms, has received little attention in social scientific research. This paper examines the effect of multiple recommender systems on different diversity dimensions. To this end, it maps different values that diversity can serve, and a respective set of criteria that characterizes a diverse information offer in this particular conception of diversity. We make use of a data set of simulated article recommendations based on actual content of one of the major Dutch broadsheet newspapers and its users (N=21,973 articles, N=500 users). We find that all of the recommendation logics under study proved to lead to a rather diverse set of recommendations that are on par with human editors and that basing recommendations on user histories can substantially increase topic diversity within a recommendation set.
Studies of informational media use have focused on individual platforms or pitted platforms against each other when investigating their connections to civic engagement. This study offers evidence, collected in a 2016 survey of US adults, of how civically engaged individuals consume various types of news content across multiple platforms. Results suggest that the best condition is the most varied one, wherein consumers get news on all six platforms studied. A breakdown of content categories finds that television viewers and those who pay attention to breaking news and crime are less civically engaged.
During the last decennia media environments and political communication systems have changed fundamentally. These changes have major ramifications for the political information environments and the extent to which they aid people in becoming informed citizens. Against this background, the purpose of this article is to review research on key changes and trends in political information environments and assess their democratic implications. We will focus on advanced postindustrial democracies and six concerns that are all closely linked to the dissemination and acquisition of political knowledge: (1) declining supply of political information, (2) declining quality of news, (3) increasing media concentration and declining diversity of news, (4) increasing fragmentation and polarization, (5) increasing relativism and (6) increasing inequality in political knowledge.
The current experimental study (N = 546) compares the effect of exposure to a television news story with a positive and negative tone on anti-immigrant attitudes and carry-over effects to uninvolved immigrant groups. Results reveal that exposure to a negatively valenced news story about North African immigrants increased negative attitudes toward that same group. Importantly, however, we find carry-over effects of exposure to a positively valenced news story about North African immigrants to attitudes toward uninvolved immigrant groups. Together these findings show that the effect of exposure to television news tone has more ramifications than previously thought, and that the impact of exposure to a positive or negative news story differs for direct effects vis-à-vis carry-over effects.
This study tests the associations between news media use and perceived political polarization, conceptualized as citizens' beliefs about partisan divides among major political parties. Relying on representative surveys in Canada, Colombia, Greece, India, Italy, Japan, South Korea, Norway, the United Kingdom, and the United States, we test whether perceived polarization is related to the use of television news, newspaper, radio news, and online news media. Data show that online news consumption is systematically and consistently related to perceived polarization, but not to attitude polarization, understood as individual attitude extremity. In contrast, the relationships between traditional media use and perceived and attitude polarization is mostly country dependent. An explanation of these findings based on exemplification is proposed and tested in an experimental design.
Over the past decade, social media platforms have penetrated deeply into the mechanics of everyday life, affecting people's informal interactions, as well as institutional structures and professional routines. Far from being neutral platforms for everyone, social media have changed the conditions and rules of social interaction. In this article, we examine the intricate dynamic between social media platforms, mass media, users, and social institutions by calling attention to social media logic—the norms, strategies, mechanisms, and economies—underpinning its dynamics. This logic will be considered in light of what has been identified as mass media logic, which has helped spread the media's powerful discourse outside its institutional boundaries. Theorizing social media logic, we identify four grounding principles—programmability, popularity, connectivity, and datafication—and argue that these principles become increasingly entangled with mass media logic. The logic of social media, rooted in these grounding principles and strategies, is gradually invading all areas of public life. Besides print news and broadcasting, it also affects law and order, social activism, politics, and so forth. Therefore, its sustaining logic and widespread dissemination deserve to be scrutinized in detail in order to better understand its impact in various domains. Concentrating on the tactics and strategies at work in social media logic, we reassess the constellation of power relationships in which social practices unfold, raising questions such as: How does social media logic modify or enhance existing mass media logic? And how is this new media logic exported beyond the boundaries of (social or mass) media proper? The underlying principles, tactics, and strategies may be relatively simple to identify, but it is much harder to map the complex connections between platforms that distribute this logic: users that employ them, technologies that drive them, economic structures that scaffold them, and institutional bodies that incorporate them.
The study addresses the question of what type of political content can trigger reactions from electoral candidates’ followers on Facebook. Citizens’ reactivity is increasingly important in contemporary political communication. The politicians’ posts can reach the wider public through the citizens’ public reactions. While we have extended knowledge about mass media reactivity, citizens’ political reactivity on social media is highly underexplored. This study is intended to fill this gap by examining what type of political content can trigger reaction from followers on politicians’ Facebook pages. The data contain 7048 Facebook posts by 183 single-member district candidates posted during the Hungarian general election campaign in 2014. The unit of analysis is the individual Facebook post, and the dependent variables are the numbers of likes, comments, and shares. The independent variables are the structural (text, picture, video, etc.) and substantial (content, emotional tone, etc.) characteristics of each post, after controlling for, inter alia, a general follower-activity score on politicians’ Facebook pages. Results showed that citizens are highly reactive to negative emotion-filled, text-using, personal, and activity-demanding posts. Virality is especially facilitated by memes, videos, negative contents and mobilizing posts, and posts containing a call for sharing.
Content-based recommender systems (CBRSs) rely on item and user descriptions (content) to build item representations and user profiles that can be effectively exploited to suggest items similar to those a target user already liked in the past. Most content-based recommender systems use textual features to represent items and user profiles, hence they suffer from the classical problems of natural language ambiguity. This chapter presents a comprehensive survey of semantic representations of items and user profiles that attempt to overcome the main problems of the simpler approaches based on keywords. We propose a classification of semantic approaches into top-down and bottom-up. The former rely on the integration of external knowledge sources, such as ontologies, encyclopedic knowledge and data from the Linked Data cloud, while the latter rely on a lightweight semantic representation based on the hypothesis that the meaning of words depends on their use in large corpora of textual documents. The chapter shows how to make recommender systems aware of semantics to realize a new generation of content-based recommenders.
We use the digitized Congressional Record and the Google Ngrams corpus to study the polarization of political discourse and the diffusion of political language since 1873. We statistically identify highly partisan phrases from the Congressional Record and then use these to impute partisanship and political polarization to the Google Books corpus between 1873 and 2000. We find that although political discourse expressed in books did become more polarized in the late 1990s, polarization remained low relative to the late 19th and much of the 20th century. We also find that polarization of discourse in books predicts legislative gridlock, but polarization of congressional language does not. Using a dynamic panel data set of phrases, we find that polarized phrases increase in frequency in Google Books before their use increases in congressional speech. Our evidence is consistent with an autonomous effect of elite discourse on congressional speech and legislative gridlock, but this effect is not large enough to drive the recent increase in congressional polarization.
In the last sixteen years, more than 200 research articles were published about research-paper recommender systems. We reviewed these articles and present some descriptive statistics in this paper, as well as a discussion about the major advancements and shortcomings and an overview of the most common recommendation concepts and approaches. We found that more than half of the recommendation approaches applied content-based filtering (55%). Collaborative filtering was applied by only 18% of the reviewed approaches, and graph-based recommendations by 16%. Other recommendation concepts included stereotyping, item-centric recommendations, and hybrid recommendations. The content-based filtering approaches mainly utilized papers that the users had authored, tagged, browsed, or downloaded. TF-IDF was the most frequently applied weighting scheme. In addition to simple terms, n-grams, topics, and citations were utilized to model users’ information needs. Our review revealed some shortcomings of the current research. First, it remains unclear which recommendation concepts and approaches are the most promising. For instance, researchers reported different results on the performance of content-based and collaborative filtering. Sometimes content-based filtering performed better than collaborative filtering and sometimes it performed worse. We identified three potential reasons for the ambiguity of the results. A) Several evaluations had limitations. They were based on strongly pruned datasets, few participants in user studies, or did not use appropriate baselines. B) Some authors provided little information about their algorithms, which makes it difficult to re-implement the approaches. Consequently, researchers use different implementations of the same recommendations approaches, which might lead to variations in the results. C) We speculated that minor variations in datasets, algorithms, or user populations inevitably lead to strong variations in the performance of the approaches. Hence, finding the most promising approaches is a challenge. As a second limitation, we noted that many authors neglected to take into account factors other than accuracy, for example overall user satisfaction. In addition, most approaches (81%) neglected the user-modeling process and did not infer information automatically but let users provide keywords, text snippets, or a single paper as input. Information on runtime was provided for 10% of the approaches. Finally, few research papers had an impact on research-paper recommender systems in practice. We also identified a lack of authority and long-term research interest in the field: 73% of the authors published no more than one paper on research-paper recommender systems, and there was little cooperation among different co-author groups. We concluded that several actions could improve the research landscape: developing a common evaluation framework, agreement on the information to include in research papers, a stronger focus on non-accuracy aspects and user modeling, a platform for researchers to exchange information, and an open-source framework that bundles the available recommendation approaches.
This paper challenges the generally taken-for-granted automatic link between media platforms, media technology and news user practices. It explores what has changed in people’s news consumption by comparing patterns in news use between 2004–2005 and 2011–2014. While new, social and mobile media technologies did not unleash a revolution in people’s dealings with news, they have facilitated, deepened and broadened user practices we already found in 2004–2005: monitoring, checking, snacking, scanning, watching, viewing, reading, listening, searching and clicking. In addition, these forms of news usage appear to increasingly order, control, organize and anchor other practices and the experience of time and environment in which they occur. Meanwhile, new and mobile news practices like linking, sharing, liking, recommending, commenting and voting have not become as central to news consumption as often assumed.
Recommender systems represent user preferences for the purpose of suggesting items to purchase or examine. They have become fundamental applications in electronic commerce and information access, providing suggestions that effectively prune large information spaces so that users are directed toward those items that best meet their needs and preferences. A variety of techniques have been proposed for performing recommendation, including content-based, collaborative, knowledge-based and other techniques. To improve performance, these methods have sometimes been combined in hybrid recommenders. This paper surveys the landscape of actual and possible hybrid recommenders, and introduces a novel hybrid, EntreeC, a system that combines knowledge-based recommendation and collaborative filtering to recommend restaurants. Further, we show that semantic ratings obtained from the knowledge-based part of the system enhance the effectiveness of collaborative filtering.
As a new communication paradigm, social media has promoted information dissemination in social networks. Previous research has identified several content-related features as well as user and network characteristics that may drive information diffusion. However, little research has focused on the relationship between emotions and information diffusion in a social media setting. In this paper, we examine whether sentiment occurring in social media content is associated with a user's information sharing behavior. We carry out our research in the context of political communication on Twitter. Based on two data sets of more than 165,000 tweets in total, we find that emotionally charged Twitter messages tend to be retweeted more often and more quickly compared to neutral ones. As a practical implication, companies should pay more attention to the analysis of sentiment related to their brands and products in social media communication as well as in designing advertising content that triggers emotions.
This experiment examines the effect of tabloid and standard packaging styles on calm and arousing news stories. The goal of this line of research is to investigate the combined influence of form and content on information processing and viewer evaluations of television news. Results indicate that the bells and whistles of tabloid production features enhance memory for calm news items but overburden the information processing system when applied to arousing news content. The evaluative measures produced data that show formal features have an influence on the meaning viewers derive from news content and that they rate news packaged in the tabloid format as less objective and believable than stories without these dramatic features.
The current debate over the extent of polarization in the American mass public focuses on the extent to which partisans' policy preferences have moved. Whereas "maximalists" claim that partisans' views on policies have become more extreme over time (Abramowitz 2010), "minimalists" (Fiorina and Abrams 2009) contend that the majority of Americans remain centrist, and that what little centrifugal movement has occurred reflects sorting, i.e., the increased association between partisanship and ideology. We argue in favor of an alternative definition of polarization, based on the classic concept of social distance (Bogardus 1947). Using data from a variety of sources, we demonstrate that both Republicans and Democrats increasingly dislike, even loathe, their opponents. We also find that partisan affect is inconsistently (and perhaps artifactually) founded in policy attitudes. The more plausible account lies in the nature of political campaigns; exposure to messages attacking the out-group reinforces partisans' biased views of their opponents.
The heuristic-systematic model (HSM) provides a general theory of social information processing. It features two modes of social information processing, a relatively effortless, top-down heuristic mode and a more effortful, bottom-up systematic mode. The model assumes that social perceivers strike a balance between effort minimization and achieving confidence in their social judgments. The HSM emphasizes three broad motivational forces: accuracy, defence, and impression motivation. Both heuristic and systematic processing can serve either of the three motives and are capable of co-occurring in an additive or interactive fashion under specified conditions. In this chapter, we describe the HSM and present illustrative research based on the model in the areas of mood and persuasion as well as minority influence.
Recommender systems support users in identifying products and services in e-commerce and other information-rich environments. Recommendation problems have a long history as a successful AI application area, with substantial interest beginning in the mid-1990s, and increasing with the subsequent rise of e-commerce. Recommender systems research long focused on recommending only simple products such as movies or books; constraint-based recommendation now receives increasing attention due to the capability of recommending complex products and services. In this paper, we first introduce a taxonomy of recommendation knowledge sources and algorithmic approaches. We then go on to discuss the most prevalent techniques of constraint-based recommendation and outline open research issues.
This study investigates the overall topic compositions and traces of right-wing populist partisanship in the content structures of German-language digital alternative news outlets. We argue that a topically heterogeneous landscape of right-wing populist alternative news outlets could maximize the reach of right-wing populist messages. Using LDA topic modeling, we explore the topical structures of nine different outlets (n = 66,965 articles) and empirically identify meaningful groups of outlets. Our analyses suggest the existence of two topically heterogeneous clusters of alternative news media: (1) core right-wing populist outlets, with around 45–50% of articles coming from topics that are related to right-wing or populist politics; and (2) topically diverse outlets. The latter group's topic selection also reveals right-wing populist partisanship, but less prominently so, while putting more emphasis on general interest news topics, such as sports, economy, or international relations. This group also includes Russian state-owned RT deutsch and Falun Gong-affiliated Epoch Times whose institutional roots, however, are only partially reflected in their topic selection. Additionally, longitudinal analyses reveal an increase of populism-related topics in both groups around the election date of the 2017 German federal election.
This work analyzes the prevalence of words denoting prejudice in 27 million news and opinion articles written between 1970 and 2019 and published in 47 of the most popular news media outlets in the United States. Our results show that the frequency of words that denote specific prejudice types related to ethnicity, gender, sexual, and religious orientation has markedly increased within the 2010–2019 decade across most news media outlets. This phenomenon starts prior to, but appears to accelerate after, 2015. The frequency of prejudice-denoting words in news articles is not synchronous across all outlets, with the yearly prevalence of such words in some influential news media outlets being predictive of those words’ usage frequency in other outlets the following year. Increasing prevalence of prejudice-denoting words in news media discourse is often substantially correlated with U.S. public opinion survey data on growing perceptions of minorities’ mistreatment. Granger tests suggest that the prevalence of prejudice-denoting terms in news outlets might be predictive of shifts in public perceptions of prejudice severity in society for some, but not all, types of prejudice.
Does the consumption of ideologically congruent news on social media exacerbate polarization? I estimate the effects of social media news exposure by conducting a large field experiment randomly offering participants subscriptions to conservative or liberal news outlets on Facebook. I collect data on the causal chain of media effects: subscriptions to outlets, exposure to news on Facebook, visits to online news sites, and sharing of posts, as well as changes in political opinions and attitudes. Four main findings emerge. First, random variation in exposure to news on social media substantially affects the slant of news sites that individuals visit. Second, exposure to counter-attitudinal news decreases negative attitudes toward the opposing political party. Third, in contrast to the effect on attitudes, I find no evidence that the political leanings of news outlets affect political opinions. Fourth, Facebook’s algorithm is less likely to supply individuals with posts from counter-attitudinal outlets, conditional on individuals subscribing to them. Together, the results suggest that social media algorithms may limit exposure to counter-attitudinal news and thus increase polarization. (JEL C93, D72, L82)
This study adopted a sentiment word database to extract sentiment-related data from microblog posts. These data were then used to investigate the effect of different types of sentiment-related words on product recommendations. The results indicate that posts containing strong sentiments received more clicks than posts containing neutral sentiments. Posts containing more than one positive sentiment word generate more effective recommendations than posts containing only one positive sentiment word. This study also demonstrated that posts with a negative polarity classification received more clicks than those with a positive polarity classification. Additionally, the microblog posts containing implicit sentiment words received more clicks than those containing explicit sentiment words. The findings presented here could assist product or service marketers who use Plurk or similar microblogging platforms better focus their limited financial resources on potential online customers to achieve maximum sale revenue.
This article is a systematic large-scale study of the reasons driving political fragmentation on social media. Making use of a comparative dataset of the Twitter discussion activities of 115 political groups in 26 countries, it shows that groups that are further apart in ideological terms interact less, and that groups that sit at the extremes of the ideological scale are particularly likely to have lower patterns of interaction. Indeed, exchanges between centrists who sit on different sides of the left–right divide are more likely than connections between centrists and extremists who are from the same ideological wing. In light of the results, theory about exposure to different ideological viewpoints online is enhanced.
More and more people read the news online, e.g., by visiting the websites of their favorite newspapers or by navigating the sites of news aggregators. However, the abundance of news information that is published online every day through different channels can make it challenging for readers to locate the content they are interested in. The goal of News Recommender Systems (NRS) is to make reading suggestions to users in a personalized way. Due to their practical relevance, a variety of technical approaches to build such systems have been proposed over the last two decades. In this work, we review the state-of-the-art of designing and evaluating news recommender systems over the last ten years. One main goal of the work is to analyze which particular challenges of news recommendation (e.g., short item life times and recency aspects) have been well explored and which areas still require more work. Furthermore, in contrast to previous surveys, the paper specifically discusses methodological questions and today's academic practice of evaluating and comparing different algorithmic news recommendation approaches based on accuracy measures.
This article argues that a common pattern and set of dynamics characterizes severe political and societal polarization in different contexts around the world, with pernicious consequences for democracy. Moving beyond the conventional conceptualization of polarization as ideological distance between political parties and candidates, we offer a conceptualization of polarization highlighting its inherently relational nature and its instrumental political use. Polarization is a process whereby the normal multiplicity of differences in a society increasingly align along a single dimension and people increasingly perceive and describe politics and society in terms of “Us” versus “Them.” The politics and discourse of opposition and the social–psychological intergroup conflict dynamics produced by this alignment are a main source of the risks polarization generates for democracy, although we recognize that it can also produce opportunities for democracy. We argue that contemporary examples of polarization follow a frequent pattern whereby polarization is activated when major groups in society mobilize politically to achieve fundamental changes in structures, institutions, and power relations. Hence, newly constructed cleavages are appearing that underlie polarization and are not easily measured with the conventional Left–Right ideological scale. We identify three possible negative outcomes for democracy—“gridlock and careening,” “democratic erosion or collapse under new elites and dominant groups,” and “democratic erosion or collapse with old elites and dominant groups,” and one possible positive outcome—“reformed democracy.” Drawing on literature in psychology and political science, the article posits a set of causal mechanisms linking polarization to harm to democracy and illustrates the common patterns and pernicious consequences for democracy in four country cases: varying warning signs of democratic erosion in Hungary and the United States, and growing authoritarianism in Turkey and Venezuela.
Democratic and Republican partisans dislike the opposing party and its leaders far more than in the past. However, recent studies have argued that the rise of affective polarization in the electorate does not reflect growing policy or ideological differences between supporters of the two parties. According to this view, though Democratic and Republican elites are sharply divided along ideological lines, differences between the policy preferences of rank-and-file partisans remain modest. In this article, we show that there is a close connection between ideological and affective polarization. We present evidence from American National Election Studies surveys that opinions on social welfare issues have become increasingly consistent and divided along party lines and that social welfare ideology is now strongly related to feelings about the opposing party and its leaders. In addition, we present results from a survey experiment showing that ideological distance strongly influences feelings toward opposing party candidates and the party as a whole.
We investigated whether there is a causal relationship between the presence of customizability technology (i.e., technology that allows individuals/websites to tailor the information environment according to user's preferences) and political selective exposure. We found that various forms of customizability technology (especially, system-driven customizability) increase selective exposure in the context of online political news consumption. Moreover, customizability technology has a stronger effect on minimizing exposure to counter-attitudinal information than it has on increasing exposure to pro-attitudinal information. The effect of customizability on selective exposure was particularly strong for ideologically moderate individuals. This study extends the understanding of the selective exposure process in today's communication environment and clarifies the implications of the Internet for deliberative democracy theory.
Online publishing, social networks, and web search have dramatically lowered the costs of producing, distributing, and discovering
news articles. Some scholars argue that such technological changes increase exposure to diverse perspectives, while others
worry that they increase ideological segregation. We address the issue by examining web-browsing histories for 50,000 US-located
users who regularly read online news. We find that social networks and search engines are associated with an increase in the
mean ideological distance between individuals. However, somewhat counterintuitively, these same channels also are associated
with an increase in an individual’s exposure to material from his or her less preferred side of the political spectrum. Finally,
the vast majority of online news consumption is accounted for by individuals simply visiting the home pages of their favorite,
typically mainstream, news outlets, tempering the consequences—both positive and negative—of recent technological changes.
We thus uncover evidence for both sides of the debate, while also finding that the magnitude of the effects is relatively
In this study, we investigated the impact of personalized news web portals on selective exposure. Results from analyses of secondary survey data from national random samples of U.S. adults show a positive relationship between personalized news and increased exposure to offline news. Users of personalized news report viewing more sources and categories of news online compared with nonusers. Partisan users of personalized news do not report increased partisan news exposure. No difference in preferences for perspective sharing or challenging news sources is found between personalized news users and nonusers. The implications for future research on personalized information systems and selective exposure are discussed.
Employing a national probability survey in 2012, this study tests relationships between social media, social network service (SNS) network heterogeneity, and opinion polarization. The results show that the use of social media is a positive predictor of the level of network heterogeneity on SNSs and that the relationship is mediated by several news-related activities, such as getting news, news posting, and talking about politics on SNSs. Testing the association between SNS network heterogeneity and polarization, this study considers 3 different dimensions of opinion polarization: partisan, ideological, and issue. The findings indicate that political discussion moderates the relationship between network heterogeneity and the level of partisan and ideological polarizations. The implications of this study are discussed.
Eli Pariser coined the term 'filter bubble' to describe the potential for online personalization to effectively isolate people from a diversity of viewpoints or content. Online recommender systems - built on algorithms that attempt to predict which items users will most enjoy consuming - are one family of technologies that potentially suffers from this effect. Because recommender systems have become so prevalent, it is important to investigate their impact on users in these terms. This paper examines the longitudinal impacts of a collaborative filtering-based recommender system on users. To the best of our knowledge, it is the first paper to measure the filter bubble effect in terms of content diversity at the individual level. We contribute a novel metric to measure content diversity based on information encoded in user-generated tags, and we present a new set of methods to examine the temporal effect of recommender systems on the user experience. We do find that recommender systems expose users to a slightly narrowing set of items over time. However, we also see evidence that users who actually consume the items recommended to them experience lessened narrowing effects and rate items more positively.
Many observers have asserted with little evidence that. Americans' social opinions have become polarized. Using General Social Survey and National Election Survey social attitude items that have been repeated regularly over 20 years, the authors ask (1) Have Americans' opinions become more dispersed (higher variance)? (2) Have distributions become flatter or more bimodal (declining kurtosis)? (3) Have opinions become more ideologically constrained within and across opinion domains? (4) Have paired social groups become more different in their opinions? The authors find little evidence of polarization over the past two decades, with attitudes toward abortion and opinion differences between Republican and Democratic party identifiers the exceptional cases.
We propose a model of motivated skepticism that helps explain when and why citizens are biased-information processors. Two experimental studies explore how citizens evaluate arguments about affirmative action and gun control, finding strong evidence of a prior attitude effect such that attitudinally congruent arguments are evaluated as stronger than attitudinally incongruent arguments. When reading pro and con arguments, participants (Ps) counterargue the contrary arguments and uncritically accept supporting arguments, evidence of a disconfirmation bias. We also find a confirmation bias—the seeking out of confirmatory evidence—when Ps are free to self-select the source of the arguments they read. Both the confirmation and disconfirmation biases lead to attitude polarization—the strengthening of t2 over t1 attitudes—especially among those with the strongest priors and highest levels of political sophistication. We conclude with a discussion of the normative implications of these findings for rational behavior in a democracy.
Balance is a notoriously difficult concept to operationalize. It has typically been investigated by examining the issues raised in elections, as well as the volume and favorability of coverage of political actors. However, even after collecting these measures, it is difficult to determine precisely what would constitute ‘balanced’ coverage. Based on a comprehensive overview of previous research in western democracies, we argue that political balance can be defined according to a political system perspective (where coverage reflects politically defined norms or regulation) or a media routine perspective (where coverage results from journalistic norms). Unless forced to follow norms, western broadcasting seems to comply with a media routine perspective. Empirically, newspaper coverage is sometimes imbalanced according to both perspectives. Finally, we discuss why only a systematic analysis of explanations across time and space makes it possible to determine whether politically ‘imbalanced’ news is the result of partisan bias or not.