Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
Poster Part (II)
Balancing Twitter data with survey information to predict electoral outcomes
Ca' Foscari University of Venice, Italy
Relevance and research question: In recent years social networks have increasingly been used to study political opinion formation, monitor electoral campaigns and predict electoral outcomes. Still, the main problem of these studies is that data from social networks are usually not a representative sample of the whole population as, for example, people using social media are generally young. In this paper we contribute to overcome sample biases by balancing Twitter data with information from a traditional surveys with the aim of nowcasting and predicting the outcome of a constitutional referendum that recently took place in Italy.
Methods & data: Data used in this research are collected from two different sources. First, using the Twitter API we collect tweets expressing voting intentions during the four weeks before the elections obtaining approximately one million tweets. Second, we use data from a traditional survey containing people’s voting intentions and demographic information such as gender and year of birth. On the first set of data we perform a sentiment analysis as proposed by Hopkins and King (2010) to study voting intentions of Twitter users. Then, to improve the social media forecast, we derive an appropriate set of weights based on the survey’ s information providing an efficient approach to balance the sample’s characteristics and adjust the forecast.
Moreover we perform a topic modelling analysis using a Latent Dirichelet Allocation model (LDA) to extract frequent topics and keywords from the Twitter data.
Results: Results show that we are correctly able to predict the outcome of the referendum, also in comparison with predictions achieved by using data from social media and traditional surveys separately.
Moreover we find that connected to voting “yes” at referendum there are positive words such as future and change while connected to voting “no” there are words such ad fear and risk.
Added value: The comparative advantage of our study is that combining data from social media with traditional survey data allows to net out most of the problems related to sample selection bias commonly present when analysing only on-line sources.
Coordination of Gender Equality Policy Resources on Twitter - A Specific Case for Political Communication
Gesis - Leibniz Institute for the Social Sciences, Germany
Relevance & Research Question:
The FP7-funded project GenPORT has initiated a user-generated online repository for resources on gender and science along with a networking space for gender equality stakeholders. The platform is amplified by the Twitter account @genderportal. The proposed presentation discusses @genderportal as a tool for the coordination of gender equality policies in science. It displays the main thematic areas that policy stakeholders are engaged with on Twitter by means of qualitative content analysis. Thereby it contributes to the transdisciplinary and constantly developing body of work on Twitter research from a qualitative perspective (Hunsinger 2005; Weller & Kinder-Kurlanda 2016; Zimmer & Proferes 2014).
Methods & Data:
Data from @genderportal were gathered with a Twitter analytics tool for four months in 2016. Subsequently, 350 tweets out of 2.509 to date, which make up 13.96% of all tweets on @genderportal, were analyzed using qualitative content analysis (Kuckartz 2012; Mayring 2007; Schreier 2012) . Tweets were clustered into categories on the basis of multiple assignments. Data on languages used, countries of audience, and the average scope of followers were gathered, thereby giving insights into user demographics of up to 596 followers at the end of the data collection process (to date @genderportal has 907 followers).
Results of the analysis display, that @genderportal followers use Twitter mainly to coordinate policy resources and exchange information about events. 159 of 320 tweets were assigned to the category “Policy”. Our findings indicate that @genderportal mainly serves to distribute current and time-sensitive information relevant to policy-makers and researchers of the gender and science community, followed by information about developments in research and policy-making.
The research conducted explores user patterns of policy stakeholders and researchers from the field of gender and science specifically. It shows that Twitter holds the potential to initiate policy learning for an international community of experts.
How does Twitter connect politicians with each other? A social network analysis of the Finnish parliament members
University of Turku, Finland
Relevance & Research Question: The social media has become an essential part of the daily politics across Europe. In Finland, for example, about 80 percent of the parliament members (MPs) utilize Twitter for networking purposes and also for spreading their political opinions. However, little research has focused on the social circles manifested in politicians’ social media networks. Previous research on the politicians’ use of social media has focused merely on the nature of political communication. This paper examines what kind of the social media networks are constituted by the Finnish MPs’ mutual following. We also analyzed what kind of party cleavages can be detected when examining MP’s shared followees.
Methods & Data: For the empirical analysis, we first extracted the network connections from the MPs’ Twitter accounts using NodeXL software. Second, we combined data from national voting advice application, which include information regarding MPs' general values and opinions regarding social political and economic issues. Our unique data with a total 161 observations from all Finnish MP’s, including a comprehensive set of background variables give us an extensive look at the MPs’ social media networks. We examined politicians’ internal networks by analyzing their reciprocal following on Twitter and external networks were analyzed by MPs’ shared followees on the Twitter.
Results: According to the explanatory analysis, a majority of MPs’ interconnections were explained by party affiliation. In addition to this, however, we also found that some clusters are clearly based on matching values, rather than party affiliations. When comparing different parties, our results suggest that the members of the Green and the National Coalition (right-wing) are highly interconnected with each other, as well as the Green and The Left. At the same time, the members of the centre-right parties seem to be withdrawn from the members of the other parties.
The Key Factors of Success of Corporate Communications on Facebook: Results of a Quantitative Content Analysis
Bielefeld University of Applied Sciences, Germany
Relevance & Research Question: With 1.8 billion monthly active users Facebook shows great potential for corporate communications. Especially small and medium-sized enterprises (SMEs) are challenged by the high intensity and dynamics of communication processes. However, what does the average profile of SMEs on Facebook look alike? Which factors lead to successful corporate-communications and thus to high interaction rates?
Methods & Data: For this purpose, we developed an application that connects to the Facebook Graph API and stores returning JSON-data of requested nodes in easy accessible CSV files with separated attributes of objects. Afterwards the data of 750 Facebook web addresses of active SMEs in DACH countries was collected and analysed.
Results: The initial results can be divided into structural and content-related usage of Facebook.
92% of SMEs use a short description of their company, followed by their business address and a link to their corporate website (90.6% each). A corporate email-address was found in 73% and a long description of the company in 61.6%. 50.9% of SMEs state their opening hours, but only 18.3% give a description of their products, 9.6% general information and 9.3% the mission of the company. Surprisingly low, only 10.2% of SMEs have their account verified by Facebook; however, 73.4% use a vanity-URL. Concerning the number of likes, accumulated 50% of SMEs count up to 356 fans, 90% up to 644.
Analysed posts [n=300.000] consist of 51.4% photos, 34% links, 7.3% status updates and 4.5% videos. The average word count is 25.4. Future results will cover more content-related usage (e.g. comments) and the identification of possible success factors for communication.
Added Value: The preliminary results of the analysis highly contribute to the general understanding of the usage of Facebook by SMEs and identifies unused structural potentials. Additionally, the collected URLs of SMEs can be considered a valid sample for future research and comparisons with large-scale enterprises, (non-governmental) organisations, certain industries, etc. Finally, the already assembled dataset resulting from SMEs allows researchers to analyse more variables and formulate hypotheses with regard to interaction rates in upcoming studies, part of which will be presented.
Automated Mechanisms to Discover and Integrate Data from Web-based Tabular Collections
1RapidMiner, Germany; 2University of Mannheim, Germany
Relevance & Research Question:
Growth in data has been phenomenal in recent years. Data exists on public sources such as the web or on corporate intranet sources. The information contained in this data can be utilized for many purposes including but not limited to marketing, research, manufacturing, fact checking. Currently, there is a lack of supporting tools to discover and integrate relevant new data to existing datasets. To address this inadequacy, several research questions need to be addressed:
• How to access heterogeneous data that resides on the web
• How to discover data that strongly matches an entity-query (comprising of an existing table and attributes to be extended)
• How to incorporate integration and fusion algorithms to manage the discovered data (tabular collections) within data mining processes
Methods & Data:
In the ongoing project DS4DM (http://ds4dm.de), we developed the Data Search extension for RapidMiner (an open source and free data mining tool). The extension provides Search-Join  as an automated process to discover and integrate tabular data.
As data source, half a million Wikipedia data tables are extracted and indexed in Lucene search engine for efficient retrieval. On this, the discovery algorithms compute schema and instance level matches for the given query. The result set is a space of candidate tables. The extension empowers the data analyst to:
• Translate and fuse new data to existing tables using the provided algorithms, or
• Perform manual refinements through i) exploratory visualizations that expose (dis)similarity patterns among the vast corpus of tabular collections and ii)graphical controls to manipulate intermediate outcomes of Search-Join process in real time (e.g. remove noisy table) but prevent loss of valuable information
• A domain-independent data enrichment solution to harness web tables
• The quality of search results is evaluated on the basis of various statistical metrics such as coverage, trust, ratio, empty values, table score
• The RapidMiner platform has been extended to incorporate data discovery and integration methods into data mining processes
 Extending RapidMiner with data search and integration capabilities. Gentile, Anna Lisa, et al., International Semantic Web Conference, 2016
The role of conspiracy beliefs and privacy concerns in online banking and shopping
Justus-Liebig-Universität Gießen, Germany
Relevance & Research Question:
Today, more than 27 million Germans use online banking. Online shopping is even more common: more than 98 % of German internet users indicated that they bought at least one product online 2015.
It has been shown, that belief in conspiracy theories leads to lower levels of trust. Furthermore, research shows that trust plays an important role in predicting the commitment to online banking and the willingness to purchase items online. The variables which determine the frequency of usage of online-services like banking and shopping as well as the choice of devices used (e.g. smartphone, tablet, etc.) are largely unknown.
Our research question is, whether conspiracy beliefs, online privacy concerns, and broad personality traits have an impact on the frequency of usage and the choice of devices used.
Methods & Data:
In our study N = 452 university students completed an online questionnaire. Participants were asked to answer questions about their online behavior in general as well as specific questions concerning their usage of online banking and shopping. Afterwards, they filled out an online privacy concerns scale and the generic conspiracy beliefs scale.
Privacy concerns and conspiracy beliefs have no influence on the frequency of usage of online banking and shopping.
But Privacy concerns are able to predict, whether a person uses online banking and shopping or not.
People with high levels of privacy concerns and conspiracy beliefs tend to use less devices for online banking and shopping in general. Also, they use more safe devices like desktop pc’s and less unsafe devices like laptops, smartphones or wearables.
When it comes to online shopping and banking, trust is an important issue. In order to increase the number of customers as well as the frequency of usage, individual differences have to be taken into consideration. In particular, privacy concerns as well as conspiracy beliefs may offer insights into ways to further improve online banking and shopping that increase feelings of trust. For vendors, it is also possible to screen the personality of their customers and adjust the level of details of the website individually.
Mediatized ideologies: Relevance, credibility and ideology of the Spanish media.
1Universitat de Valencia (Spain); 2GIPEyOP; 3Mediaflows
Relevance & Research Question: The media are at the center of the political decision-making process carried out by citizens. The media concentrate information on the topics and leaders of their choice, strongly conditioning public’s perception. In southern European countries, this is combined with a news scenario in which a majority of media outlets and the major factions of political power have adopted what Hallin and Mancini (2004) call a “model of polarized pluralism”. The political and electoral decision-making processes carried out by citizens can only be understood identifying the ideological leanings of the media through which they perceive the events which comprise the public agenda.
This paper studies how the electorate perceives the major media outlets in terms of credibility and ideology. We seek to answer, among others, the following questions: which media are considered the most influential in public opinion formation in Spain; what ideological position Spaniards assign to the major media outlets and the most important journalists and public figures; what links can be established between Spaniards’ media consumption habits and their perception of politics in terms of ideology, influence and credibility.
Methods & Data: An online poll of 2,500 electors, to be performed in November-December 2016, will gather information on the 25 most important national media outlets, as well as the characteristics of the poll respondents. We will use statistical techniques, including logit models and correspondence analysis, to find the most appropriate answers to the questions posed.
Results: The results obtained will offer a broad overview of the Spanish communications panorama in terms of its ideological distribution. They will shed light on the public’s perception of the ideological positions of the major Spanish media outlets, as well as the public’s consumption criteria, in the context of the polarization that Spanish politics has been undergoing over the last few years.
Added-Value: This is the first study carried out in Spain on the electorate’s perception—politically speaking—of the relevance, credibility and ideology of the media. This will allow us to position media and voters and to determine different subsets of voters’ levels of exposure to different political messages.
How online gaming affects social capital and social support: a survey study on World of Warcraft
University of Muenster, Germany
Relevance & Research Question:
The increasing popularity of online games has raised questions about the social consequences of excessive play. While some scholars found evidence that playing may lead to isolation, others argued that online gaming offers great opportunities to build up new friendships and thus to increase players’ social capital. The current study takes both possibilities into consideration. Based on a sample of 409 players of the popular role-playing game World of Warcraft (WoW), possible positive as well as possible negative social effects of playing are investigated. To find out more precisely under which circumstances online gaming has positive or negative social consequences, not only playing time, but also specific game-related behaviors of the players and their playing motivations are considered as predictors of social outcomes of playing.
Methods & Data:
Regression analysis was used to investigate how playing time, game-related behaviors (membership in a clan, preference for team play) and motivations (socializing, escapism, and achievement) are connected with different types of social outcomes (bridging and bonding social capital received from playing; overall social support). Further, moderation analysis was performed to investigate whether players’ motivations moderated the connection between time spent playing and social outcomes of playing.
Playing WoW for social motivations and preferring team play over solo play was connected with increases in the acquisition of social capital as well as with increases in social support in general. Excessive play showed negative connections with the overall social embeddedness of the players. However, moderation analysis revealed that social motivations seem to buffer some of the negative social consequences of excessive play, while playing for escapism seems to act as an amplifier.
The results show that the social outcomes players receive from WoW are not the same for all players, but depend on their motivations and their game-related behaviors. In this, the study underlines the importance of considering additional underlying factors when investigating social consequences of playing online games. Online gaming is not a one-dimensional activity and thus solely looking at connections between time spent playing and social outcomes of playing is not sufficient.
Contact and Legal Notice · Contact Address:
Conference: GOR 17
|Conference Software -
ConfTool Pro 2.6.96
© 2001 - 2016 by H. Weinreich, Hamburg, Germany