Measuring Information Exposure in Dynamic and Dependent Networks (ExpoNet)
Lead Research Organisation:
University of Exeter
Department Name: Politics
Abstract
The advent of Web 2.0 - the second generation of the World Wide Web, that allows users to interact, collaborate, create and share information online, in virtual communities - has radically changed the media environment, the types of content the public is exposed to as well as the exposure process itself. Individuals are faced with a wider range of options (from social and traditional media), new patterns of exposure (socially mediated and selective), and alternate modes of content production (e.g. user-generated content). In order to understand change (and stability) in opinions and behaviour, it is necessary to measure to what information a person has been exposed. The measures social scientists have traditionally used to capture information exposure usually rely on self-reports of newspaper reading and television news broadcast viewing. These measures do not take into account that individuals browse and share diverse information from social and traditional media on a wide range of platforms. According to the OECD's Global Science Forum 2013 report, social scientists' inability to anticipate the Arab Spring was partly due to a failure to understand 'the new ways in which humans communicate' via social media and the ways they are exposed to information. And social media's mixed record for predicting the results of recent UK elections suggests better tools and a unified methodology are needed to analyze and extract political meaning from this new type of data.
We argue that a new set of tools, which models exposure as a network and incorporates both social and traditional media sources, is needed in the social sciences to understand media exposure and its effects in the age of digital information. Whether one is consuming the news online or producing/consuming information on social media, the fundamental dynamic of consuming public affairs news involves formation of ties between users and media content by a variety of means (e.g. browsing, social sharing, search). Online media exposure is then a process of network formation that links sources and consumers of content via their interactions, requiring a network perspective for its proper understanding. We propose a set of scalable network-oriented tools to 1) extract, analyse, and measure media content in the age of "big media data", 2) model the linkages between consumers and producers of media content in complex information networks, and 3) understand co-development of network structures with consumer attitudes/behaviours.
In order to develop and validate these tools, we bring together an interdisciplinary and international team of researchers at the interface of social science and computer science. Expertise in network analysis, text mining, statistical methods and media analysis will be combined to test innovative methodologies in three case studies including information dynamics in the 2015 British election and opinion formation on climate change. Developing a set of sophisticated network and text analysis tools is not enough, however. We also seek to build national capacity in computational methods for the analysis of online 'big' data.
We argue that a new set of tools, which models exposure as a network and incorporates both social and traditional media sources, is needed in the social sciences to understand media exposure and its effects in the age of digital information. Whether one is consuming the news online or producing/consuming information on social media, the fundamental dynamic of consuming public affairs news involves formation of ties between users and media content by a variety of means (e.g. browsing, social sharing, search). Online media exposure is then a process of network formation that links sources and consumers of content via their interactions, requiring a network perspective for its proper understanding. We propose a set of scalable network-oriented tools to 1) extract, analyse, and measure media content in the age of "big media data", 2) model the linkages between consumers and producers of media content in complex information networks, and 3) understand co-development of network structures with consumer attitudes/behaviours.
In order to develop and validate these tools, we bring together an interdisciplinary and international team of researchers at the interface of social science and computer science. Expertise in network analysis, text mining, statistical methods and media analysis will be combined to test innovative methodologies in three case studies including information dynamics in the 2015 British election and opinion formation on climate change. Developing a set of sophisticated network and text analysis tools is not enough, however. We also seek to build national capacity in computational methods for the analysis of online 'big' data.
Planned Impact
The tools developed as part of will contribute to better understand the dynamic interplay in informational flows between users, traditional news outlets and new social media and to examine the relationship between information exposure and public opinion/behaviour. The implications of this research can also inform debates about the effectiveness of media policy and regulation in influencing citizens' political beliefs and engagement as well as their environmental attitudes and behaviour. In this direction, the outcomes of ExpoNet will allow us to address key ESRC strategic priorities such as influencing behaviour and informing interventions.
Potential Policy and Societal Impact:
The outcomes of this project will benefit a large and diverse user community outside academia. Potential beneficiaries include government agencies, third sector organisations, industry organisation and practitioners interested in understanding the central characteristics of the complex information environment prevalent in modern societies and in assessing how it affects mass attitudes and behaviours. One segment of this extra-academic user group comprises policy-makers involved in the design of media policy as well as public institutions interested its societal consequences, such as the Department for Culture, Media and Sport, the Department for Education, and the Electoral Commission. A different group consists of non-governmental organisations interested in the impact of media messages on political participation, and such as International IDEA, the Institute for Government, the Electoral Society and Demos. Another user group consists of media organisations and journalists who provide audiences - the mass audience as well as more specialised and targeted audiences - with (political) information. Finally, the outputs of this project will also benefit key industry user-groups, including social research and market firms and marketing agencies.
Non-academic users and potential beneficiaries will be invited to participate in our stakeholder and dissemination workshop, where the web application developed by SeeDATA and research outputs of ExpoNet will be presented and their implications for media policy and regulation will be discussed. We will actively seek to recruit participants taking advantage of existing partnerships with organizations like Kieskompas, OfCom, and public broadcasters stemming from research networks the research team members are currently involved in. The members of the project team have worked and engaged in consulting with non-governmental and governmental organizations, and thus have experience in communicating research findings to wider audiences. We will also rely on the ESRC Communications Toolkit to reach wider non-academic communities.
Research Area Impact:
Another benefit of the research is to the UK Research Area: the visibility of the UK as a venue for research on computational social science and big data analysis. The innovative set of methodological tools and associated user-friendly software/source code developed during the life of the project to conduct and integrate statistical network and text analysis will have applicability to a rich set of questions in the social sciences in general, and will contribute to place the UK at the forefront of this growing research area.
Potential Policy and Societal Impact:
The outcomes of this project will benefit a large and diverse user community outside academia. Potential beneficiaries include government agencies, third sector organisations, industry organisation and practitioners interested in understanding the central characteristics of the complex information environment prevalent in modern societies and in assessing how it affects mass attitudes and behaviours. One segment of this extra-academic user group comprises policy-makers involved in the design of media policy as well as public institutions interested its societal consequences, such as the Department for Culture, Media and Sport, the Department for Education, and the Electoral Commission. A different group consists of non-governmental organisations interested in the impact of media messages on political participation, and such as International IDEA, the Institute for Government, the Electoral Society and Demos. Another user group consists of media organisations and journalists who provide audiences - the mass audience as well as more specialised and targeted audiences - with (political) information. Finally, the outputs of this project will also benefit key industry user-groups, including social research and market firms and marketing agencies.
Non-academic users and potential beneficiaries will be invited to participate in our stakeholder and dissemination workshop, where the web application developed by SeeDATA and research outputs of ExpoNet will be presented and their implications for media policy and regulation will be discussed. We will actively seek to recruit participants taking advantage of existing partnerships with organizations like Kieskompas, OfCom, and public broadcasters stemming from research networks the research team members are currently involved in. The members of the project team have worked and engaged in consulting with non-governmental and governmental organizations, and thus have experience in communicating research findings to wider audiences. We will also rely on the ESRC Communications Toolkit to reach wider non-academic communities.
Research Area Impact:
Another benefit of the research is to the UK Research Area: the visibility of the UK as a venue for research on computational social science and big data analysis. The innovative set of methodological tools and associated user-friendly software/source code developed during the life of the project to conduct and integrate statistical network and text analysis will have applicability to a rich set of questions in the social sciences in general, and will contribute to place the UK at the forefront of this growing research area.
Publications
Banducci S
(2020)
Exposure to campaign claims in mobile browsing and survey data
Banducci S
(2018)
Intermedia agenda setting in personalized campaigns: How news media influence the importance of leaders
in Electoral Studies
Banducci S
(2018)
A little justification goes a long way: audience costs and the EU referendum
in Acta Politica
Banducci S
(2017)
The Channels of Democracy
Banducci SA
(2016)
EU Referendum Analysis 2016
Banducci, S
(2022)
Handbook on Politics and Public Opinion
Boussalis C
(2019)
Communicating Climate Mitigation and Adaptation Efforts in American Cities
in Climate
Boussalis C
(2017)
Elite polarization and correcting misinformation in the "post-truth era".
in Journal of Applied Research in Memory and Cognition
Boussalis C
(2018)
Climate change communication from cities in the USA
in Climatic Change
Boussalis C
(2016)
Measuring and modeling Russian newspaper coverage of climate change
in Global Environmental Change
| Description | Technical and Methodological Developments One of our project objectives is to develop a set of tools for analysing digital data. In line with this, we developed a series of protocols and code that implement different methods relevant to project goals. Several of these methods are currently being developed into a series of short help sheets with attendant short (90 second) technical videos. These will deposited on NCRM page, on an EXPONet github account https://github.com/EXPONetProject, and on a research cluster webpage https://mediaeffectsresearch.wordpress.com/exponet/intro-to-jupyter-notebook/). Code developed so far includes methods for: 1) Harvesting data from traditional news (online editions), news broadcasts (transcripts), and social media (Twitter and public Facebook pages). Methods include techniques for use of cloud computing (Amazon AWS) to parallelise data collection, enabling rapid collection of large historical social media datasets. 2) Automated coding of web content (e.g. HTML pages, URLs). Substantial progress has been on code pipelines for separating the "true" content of a webpage from unwanted content such as embedded adverts. Further work has developed scripts for determining whether a given URL points to a user-facing webpage or to unwanted supporting web sources (such as hidden domains used to direct embedded advertising). 2) Extracting social networks from large volumes of Twitter data. Techniques include standard methods such as retweet/mention interactions, alongside novel methods such as using the "one hop" neighbourhood of adjacent nodes to give greater density and infer community structure in sparse networks. 3) Extracting shared links from Twitter content and constructing "shared audience" networks of web domains linked by individual users who shared content from both. These networks allow similar exposure patterns to be identified in the user population, and clusters of web domains to be identified based on the overlap in the user groups who access their content. 4)Temporal clustering of networks based on similar community structure. This can be used as a dimension reduction technique for complex media exposure networks that evolve over time. 5) Tracking cascading "reply" activity in Twitter, enabling tracking of conversations started around a single post or event. 6)Tools for linking web activity of users (taken from comprehensive activity logs provided by volunteers from the ICM panel, see below) to longitudinal survey responses. 7) We wrote an R/C++ implementation for a number of supervised, generative text classification models shown to perform well for solving large, multi-label classification problems. We are in the process of finalizing an R package (superlda) that will be made available to researchers on CRAN and GitHub. 8) We developed a custom web-scraping and crawling framework built on top of the open-source Python library scrapy (https://scrapy.org). We have deployed some of these tools in our research outputs. For example, we use individual data on browsing histories combined with survey data to examine whether online news exposure exhibits signs of segregation and selectivity. By using online news behaviour combined with survey reports of attitudes, we can capture exposure to both traditional news sources and news shared via social media platforms. Most importantly, we can also examine what types of individuals (e.g. partisans, educated) are more likely to exhibit selective tendencies. 10) We were able to compare methods for estimating ideological position when applied to media text. We tested three methods of estimating ideological bias in news media stories. This forms the basis for the development of a news reading app. We find that WordScores offers the most reliable estimate and reliability is improved when applied after identifying topic. We noted above the lack of development of testing candidate statistical network. On the other hand, we were able to innovative in the use of web browsing history data linked to survey data as a method of estimating information exposure. Substantive findings from case studies: We find, consistent with recent empirical work, the extent of segregation in exposure may be overstated. Furthermore, the degree of segregation and selectivity varies across groups that are defined by holding shared political preferences. For example, in the case of Brexit, those who supported the 'Leave' side were more selective in their news exposure. Our approach allows comparison of news exposure patterns by domains versus news exposure to topics. To our knowledge, this is the first analysis to allow this comparison (Banducci et al. 2018) Co-I Professor David Lazer published his work on the project in Science contributing to the debate on Fake New. His findings indicate: "Only 1% of individuals accounted for 80% of fake news source exposures, and 0.1% accounted for nearly 80% of fake news sources shared. Individuals most likely to engage with fake news sources were conservative leaning, older, and highly engaged with political news. A cluster of fake news sources shared overlapping audiences on the extreme right, but for people across the political spectrum, most political news exposure still came from mainstream media outlets" (Grinberg et al. 2019). Work on estimating network structures has led to the following substantive conclsions: "We find that the social network structure linking MPs and MEPs evolves over time, with distinct communities forming and re-forming, driven by party affiliations and political events. Without including any information about time in our model, we nevertheless find that the evolving social network structure shows multiple persistent and recurring states of affiliation between politicians, which align with content states derived from topic analysis of tweet text. These findings show that the dominant state of partisan segregation can be challenged by major political events, ideology, and intra-party tension that transcend party affiliations" (Weaver et al. 2018). |
| Exploitation Route | Currently, we envision that the academic community will make use of the tolls available through EXPONet through their own research. Our findings will contribute toward discussions of the impact of social media on the polarisation of attitudes in society and the electorate. They will encourage further refinement of the impact of media exposure in an online environment. |
| Sectors | Communities and Social Services/Policy,Digital/Communication/Information Technologies (including Software),Government, Democracy and Justice |
| URL | https://github.com/EXPONetProject/tools |
| Description | We see three areas of impact: 1. Career development and capacity building on computational social science: There were four PDRA who were recruited to the project: Dr Iulia Cioroianu (now permanent lecturer at Bath University), Dr Stefan Wojcik (Lecturer at University of Colorado); Dr Iain Weaver (currently Research Fellow, Exeter); Dr Danielle Martin; Dr Charlotte James (currently Research Fellow, Exeter). We have also further built capacity by engaging MA level students in our project through the use of process data for projects and dissertations, training of PhD students in computational methods and offering training through NCRM. The project has also been instrumental in developing Exeter's relationship with Turing. Banducci is a Turing fellow (2018-2020) and our final dissemination event was held at Turing. We have also delivered a successful training programme in computational methods in partnership with NCRM (follow on funding), our Institute of Data Science and AI and the Turing in 2022 (see details under engagement). This training and capacity building impact improved the computational skills of over 50 participants. 2. Informing the public debate on digital campaigning, social media and misinformation: Findings from the research project on information exposure in the new media ecosystem have been used to inform the debate on digital communication and its role in democracy. In particular our research has formed the basis of expert testimony at the House of Lords, has been referenced in a BBC televised programme and has been covered in the press (New Science and Time). 3. Developing evidence based civic technology applications to increase public understanding: -The UNBIAS chrome browser extension evaluates the ideological slant of an article that the user submits to the app. After evaluating the article based on a text classifier, the user is given the ideological slant in the article and then offered 2 other articles to read with a different ideological slant. - The UNBIAS iPhone app (available in the iPhone app store) is a platform we have developed to evaluate a range of text characteristics encountered while seeking news and information. Currently, users who have installed the app can share a news story they are interested to the app and the text evaluated for gender bias. The gender bias classifier is one of the outcomes of our protocol for annotating text and training classifiers based on the crowd and human annotated text. |
| First Year Of Impact | 2021 |
| Sector | Digital/Communication/Information Technologies (including Software),Government, Democracy and Justice |
| Impact Types | Societal,Policy & public services |
| Description | Citation in Voter engagement in the UK:Fourth Report of Session 2014-15 |
| Geographic Reach | National |
| Policy Influence Type | Citation in other policy documents |
| URL | http://www.publications.parliament.uk/pa/cm201415/cmselect/cmpolcon/232/232.pdf |
| Description | Committee on Political Polling and Digital Media (House of Lords) - Oral Evidence |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| URL | https://www.parliament.uk/business/committees/committees-a-z/lords-select/political-polling-digital-... |
| Description | Political and Constitutional Reform Committee (House of Commons)- UK Voter Engagement - Written Evidence |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| URL | http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/political-and-const... |
| Description | Political and Constitutional Reform Committee (House of Commons)- UK Voter Engagement -- Oral Evidence |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| URL | http://www.parliament.uk/business/committees/committees-a-z/commons-select/political-and-constitutio... |
| Description | Written Evidence to Commons Select Committee Inquiry on Women in the House of Commons after the 2020 Election |
| Geographic Reach | National |
| Policy Influence Type | Contribution to a national consultation/review |
| URL | https://www.parliament.uk/business/committees/committees-a-z/commons-select/women-and-equalities-com... |
| Description | (OPTED) - Observatory for Political Texts in European Democracies - A European Research Infrastructure |
| Amount | € 2,998,500 (EUR) |
| Funding ID | 951832 |
| Organisation | European Commission |
| Sector | Public |
| Country | European Union (EU) |
| Start | 10/2020 |
| End | 09/2023 |
| Description | CREST: Centre for Research and Evidence on Security Threats |
| Amount | £95,617 (GBP) |
| Funding ID | ES/N009614/1 |
| Organisation | Economic and Social Research Council |
| Sector | Public |
| Country | United Kingdom |
| Start | 09/2016 |
| End | 09/2017 |
| Description | ERC Advanced Grant - TWICEASGOOD |
| Amount | € 2,500,000 (EUR) |
| Funding ID | 101019284 |
| Organisation | European Research Council (ERC) |
| Sector | Public |
| Country | Belgium |
| Start | 01/2022 |
| End | 12/2026 |
| Description | ESRC IAA UKRI Innovation Fellowships Funding |
| Amount | £23,464 (GBP) |
| Organisation | Economic and Social Research Council |
| Sector | Public |
| Country | United Kingdom |
| Start | 10/2017 |
| End | 03/2018 |
| Description | IBM Faculty Awards |
| Amount | $8,000 (USD) |
| Organisation | IBM |
| Sector | Private |
| Country | United States |
| Start | 01/2018 |
| End | 01/2019 |
| Description | National Centre for Research Methods 2020-2024 |
| Amount | £3,181,119 (GBP) |
| Funding ID | ES/T000066/1 |
| Organisation | Economic and Social Research Council |
| Sector | Public |
| Country | United Kingdom |
| Start | 01/2020 |
| End | 12/2024 |
| Description | Populism's Roots: Economic and Cultural Explanations in Democracies of Europe (PRECEDE) |
| Amount | € 998,700 (EUR) |
| Funding ID | 96 999 |
| Organisation | Volkswagen Foundation |
| Sector | Charity/Non Profit |
| Country | Germany |
| Start | 07/2020 |
| End | 06/2023 |
| Description | Turing Fellow [Alan Turing Institute] |
| Amount | £15,000 (GBP) |
| Organisation | Alan Turing Institute |
| Sector | Academic/University |
| Country | United Kingdom |
| Start | 09/2018 |
| End | 08/2020 |
| Title | 2016 EU Referendum campaign online news and information URLs |
| Description | The data set represents processed data from individual web browsing histories collected during the EU Referendum campaign as part of ICM Unlimited Reflected Life's panel. Each line of data represents the number of times an individual user visited a news & information domain during the data collection period. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | We expect to published several research papers with the data. Furthermore, working with ICM we will develop several tools for the analysis of the data. |
| URL | https://reshare.ukdataservice.ac.uk/cgi/users/home?screen=EPrint::View&eprintid=854256 |
| Title | 2016 EU Referendum campaign online news and information URLs |
| Description | The data set represents processed data from individual web browsing histories collected during the EU Referendum campaign as part of ICM Unlimited Reflected Life's panel. Each line of data represents the number of times an individual user visited a news & information domain during the data collection period. The advent of Web 2.0 - the second generation of the World Wide Web, that allows users to interact, collaborate, create and share information online, in virtual communities - has radically changed the media environment, the types of content the public is exposed to as well as the exposure process itself. Individuals are faced with a wider range of options (from social and traditional media), new patterns of exposure (socially mediated and selective), and alternate modes of content production (e.g. user-generated content). In order to understand change (and stability) in opinions and behaviour, it is necessary to measure to what information a person has been exposed. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | The data set represents processed data from individual web browsing histories collected during the EU Referendum campaign as part of ICM Unlimited Reflected Life's panel. Each line of data represents the number of times an individual user visited a news & information domain during the data collection period.The dataset has been used to understand news exposure online. |
| URL | https://beta.ukdataservice.ac.uk/datacatalogue/studies/study?id=854256 |
| Title | Brexit Panel Survey |
| Description | Survey data from a 3 wave panel with field dates in February, April and June in 2016. Respondents are drawn from ICM Research's online panel. Questions focused on media use, remain/leave opinions and other attitudinal measures. The sample and participation in waves is given in below tabl wave_presence | Freq. Percent Cum. ------------+----------------------------------- A | 247 21.40 21.40 AJ | 69 5.98 27.38 F | 291 25.22 52.60 FA | 166 14.38 66.98 FAJ | 109 9.45 76.43 FJ | 42 3.64 80.07 J | 230 19.93 100.00 ------------+----------------------------------- Total | 1,154 100.00 |
| Type Of Material | Database/Collection of data |
| Provided To Others? | No |
| Impact | We expect to publish several research papers with the data. We expect to contribute to debates on social media, filter bubbles and misinformation in campaigns based on the results of this research. |
| Title | Brexit Twitter data |
| Description | Database of Tweets collected during the 2016 UK referendum on EU membership. The tweets were collected through the REST API, using keywords associated with politics and the referendum. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2017 |
| Provided To Others? | No |
| Impact | The data was used for two working papers, one presentation at the 2017 NCRM Autumn School, and one NCRM Online Resources video podcast. |
| Title | Collecting tweets and constructing a retweet network |
| Description | Methods for collecting Twitter data (public tweets, detailed account information, retweets, etc.), constructing a network of retweets, and detecting communities of well inter-connected individuals. |
| Type Of Material | Data analysis technique |
| Year Produced | 2017 |
| Provided To Others? | Yes |
| Impact | The methods were used to analyze the evolving community structure of UK Members of Parliament and Members of the European Parliament in the following paper (under review): Iain S. Weaver, Hywel Williams, Iulia Cioroianu, Susan Banducci, Travis Coan. 2017. "Dynamic Social Media Affiliations Among UK Politicians" |
| URL | https://github.com/EXPONetProject/tools |
| Title | Data for Ideological biases in social sharing of online information about climate change |
| Description | This repository contains an anonymised dataset to support the paper "Ideological biases in social sharing of online information about climate change" by Tristan J.B. Cann, Iain S. Weaver and Hywel T.P. Williams, submitted for publication in PLOS ONE.The files present contain the following:tweet_ids - A list of all tweets ids used in the study.coded_urls - A list of the (up to) five most common URLs from each of the 75 most common domains. Where these were not social media sites and content was available, they were graded for political and climate bias by the human coders.domain_bias_grades - A list of domains and the final bias scores assigned to them following the standardisation process we applied to the scores received from our coders. The first line of this file is a header labelling the four columns as political bias, climate change bias, political standard deviation and climate change deviation.The networks folder contains subfolders for each of the seven weeks studied. Three files are provided for each week.week_x_bipartite_edges - A list of source, target pairs to define edges in the bipartite user-URL network. Source and target give the user and URL node IDs respectively. Pairs are not guaranteed to be unique, and duplicates should increment the edge weight.week_x_url_labels - A list of expanded URLs given in the order corresponding to the edge list described above.week_x_user_labels - A list of anonymised user IDs given in the order corresponding to the edge list for this week. These anonymised numeric user identifiers are consitent across each week for cross referencing. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| URL | https://figshare.com/articles/dataset/Data_for_Ideological_biases_in_social_sharing_of_online_inform... |
| Title | Data for Ideological biases in social sharing of online information about climate change |
| Description | This repository contains an anonymised dataset to support the paper "Ideological biases in social sharing of online information about climate change" by Tristan J.B. Cann, Iain S. Weaver and Hywel T.P. Williams, submitted for publication in PLOS ONE.The files present contain the following:tweet_ids - A list of all tweets ids used in the study.coded_urls - A list of the (up to) five most common URLs from each of the 75 most common domains. Where these were not social media sites and content was available, they were graded for political and climate bias by the human coders.domain_bias_grades - A list of domains and the final bias scores assigned to them following the standardisation process we applied to the scores received from our coders. The first line of this file is a header labelling the four columns as political bias, climate change bias, political standard deviation and climate change deviation.The networks folder contains subfolders for each of the seven weeks studied. Three files are provided for each week.week_x_bipartite_edges - A list of source, target pairs to define edges in the bipartite user-URL network. Source and target give the user and URL node IDs respectively. Pairs are not guaranteed to be unique, and duplicates should increment the edge weight.week_x_url_labels - A list of expanded URLs given in the order corresponding to the edge list described above.week_x_user_labels - A list of anonymised user IDs given in the order corresponding to the edge list for this week. These anonymised numeric user identifiers are consitent across each week for cross referencing. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2021 |
| Provided To Others? | Yes |
| URL | https://figshare.com/articles/dataset/Data_for_Ideological_biases_in_social_sharing_of_online_inform... |
| Title | From URLs to Topic Models |
| Description | The script processes URLs from social media data or web browsing histories, extracts the text and title, and identifies topics in the article text using a Latent Dirichlet Allocation model. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2017 |
| Provided To Others? | Yes |
| Impact | The models were made available to other researchers on GitHub and were presented at the 2017 NCRM Autumn School in Southampton. A video podcast was recorded for the NCRM Online Resources series. |
| URL | https://github.com/EXPONetProject/tools/blob/master/From%20URLs%20to%20Topic%20Models.ipynb |
| Title | Social media bias, trust and practices 2019 |
| Description | In order to develop appropriate tools (e.g. a mobile app) we explored through a participant survey the issues such as the kinds of media coverage that engage and inform voters, whether and how this varies by subgroups such as generation, and the aspects of campaigns that contribute to more positive views of the political process. As part of ExpoNet's objectives to understand news and information exposure in the contemporary environment, we worked to to enhancing the quality of representative democracy through giving better access to citizens to quality information and the tools necessary to evaluate the news they consumed. By providing information about the nature and quality of traditional and new media election coverage over time and its impact on individuals, our research will offer pointers towards how to mobilize informed engagement with campaigns and in elections. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2020 |
| Provided To Others? | Yes |
| Impact | Publications,reports should result from sharing of database. |
| URL | https://reshare.ukdataservice.ac.uk/cgi/users/home?screen=EPrint%3A%3AView&eprintid=854260 |
| Description | Clicking on Brexit |
| Organisation | ICM Research |
| Country | United Kingdom |
| Sector | Private |
| PI Contribution | Working with ICM we designed a survey that is connected to individual respondent's clickstream data. We have created tools for analysing the data. |
| Collaborator Contribution | ICM field the survey and collected the clickstream data. |
| Impact | The data collected was briefly reported in this publication: Susan Banducci and Dan Stevens. 2016. "Myth versus fact: are we living in a post-factual democracy?" in Jackson, D., Thorsen, E. and Wring, D.,(eds) EU Referendum Analysis 2016: Media, Voters and the Campaign. Discussion Paper. Bournemouth, UK: The Centre for the Study of Journalism, Culture and Community. http://www.referendumanalysis.eu/ |
| Start Year | 2016 |
| Description | UNBias - Understanding News Information Bias |
| Organisation | IBM |
| Country | United States |
| Sector | Private |
| PI Contribution | Database of articles from UK news sources, processed and coded on CrowdFlower. Dr. Cioroianu's work visits to Amsterdam (3 visits and remote work). |
| Collaborator Contribution | Funded and supervised a Computer Science Master's student working on the project. Working space for Dr. Cioroianu's work visits. Sponsored Dr. Cioroianu's application for an IBM Faculty Award. |
| Impact | Database of articles coded by topic and ideology. Computer algorithms and scripts for processing news articles, extracting topic and ideology. Working paper on methods for extracting topics and ideology from texts. |
| Start Year | 2017 |
| Title | UNBIAS - a chrome browser extension |
| Description | The UNBIAS app evaluates the ideological slant of an article that the user submits to the app. After evaluating the article based on a text classifier, the user is given the ideological slant in the article and then offered 2 other articles to read with a different ideological slant. |
| Type Of Technology | Webtool/Application |
| Year Produced | 2018 |
| Impact | There are no impacts yet as we are still in the pilot stage. |
| Description | ? Presentation to IBM Centre for Advanced Study (CAS) & IBM Europe, |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | We presented our work on social networks during the Brexit campaign using tools developed in the project and Twitter data. Presentation was by Williams & Banducci Nov 2016]. Following on from this presentation, we were invited to the CAS centre in Amsterdam. We are now examining ways to collaborate on EXPONet linked research and training activities. |
| Year(s) Of Engagement Activity | 2016 |
| Description | Analysis for News Website on Romanian Elections |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | ? We have used the methods developed for Facebook data collection to analyze key elements in the structure of political communication around the November 2016 Romanian elections. The analysis was published by two major Romanian news websites. http://www.contributors.ro/politica-doctrine/campania-electorala-pe-facebook-11-30-noiembrie-2016/ http://www.hotnews.ro/stiri-opinii-21451250-campania-electorala-facebook-11-30-noiembrie-2016.htm |
| Year(s) Of Engagement Activity | 2016 |
| URL | http://www.contributors.ro/politica-doctrine/campania-electorala-pe-facebook-11-30-noiembrie-2016/ |
| Description | Blog post and report on corruption in Fiji |
| Form Of Engagement Activity | Engagement focused website, blog or social media channel |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Policymakers/politicians |
| Results and Impact | We wrote a report and accompanying blog post for Fiji's Independent Commission Against Corruption (FICAC). FICAC now receives over 10,000 complaints each year about a range of issues from people across the country concerned about corruption, mismanagement, and abuse of power. Our report analyses all complaints made to FICAC between 2007 and 2014 (drawing on data not publicly available), and provide a detailed long term view on the type of these complaints. |
| Year(s) Of Engagement Activity | 2017,2018 |
| URL | http://devpolicy.org/key-trends-from-complaints-to-fijis-independent-commission-against-corruption-2... |
| Description | Co-organising the Computational Social Science Specialist Group for the Political Studies association |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Other audiences |
| Results and Impact | I have submitted an application to co-organise the Computational Social Science Specialist Group for the Political Studies Association. The goal of the group is to promote Computational Social Science in the UK, as well as provide a forum for discussing and collaborating on CSS-related projects. |
| Year(s) Of Engagement Activity | 2018 |
| Description | Co-organizer of the R Users Group -- Devon |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Postgraduate students |
| Results and Impact | I co-organise the R Users Group, Devon, which hosts local meetups between University staff, students, and regional industry to share insights on the R programming language. |
| Year(s) Of Engagement Activity | 2017,2018 |
| Description | Computational Communication Science Spring School |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | NCRM UoE Computational Communication Methods Spring School Researchers interested in computational social science will be given the chance to learn new skills at a spring school in April 2022. The NCRM/Exeter Computational Communication Methods Spring School will provide training at introductory and advanced levels, catering for both social scientists and data scientists. The school will take place at the University of Exeter over two 4-day sessions on 6-16 April 2022 and is co-sponsored by:- IDSAI Computational Social Science Social Data Science Group, Turing Institute Exeter Q-Step The programme will cover multiple computational approaches, such as machine learning and network analysis, and their application to communication research looking at text, images and social media data. World-leading experts will deliver workshops, seminars and demonstrations, help desks will offer one-to-one consultations and there will be opportunities for more informal networking. Programme Overview Week 1: Text Analysis & Machine Learning with AI (4 day introductory session) Week 2: Advanced Sessions (Four 1-day sessions). Topics will include: Social Media Networks (Monday) delivered by Dr Iulia Cioroianu Deep Learning and AI (Tuesday) delivered by Dr Pikakshi Manchanda Data Ethics for Computational Research (Wednesday) delivered by Dr Silvia Milano Image Analysis (Thursday) delivered by Dr Constantine Boussalis |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://www.exeter.ac.uk/news/events/details/index.php?event=11773 |
| Description | Computational social science |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | Presentation at NCRM ESRC research Methods Festival by Professor David Lazer The advent of Web 2.0 has radically changed the media environment, the types of content the public is exposed to as well as the exposure process itself. There is a wider range of options (from social and traditional media), new patterns of exposure (socially mediated and selective), and alternate modes of content production (e.g. user-generated content). The session introduces, EXPONet is a 3-year ESRC/NCRM project that bridges recent advances in the literature on statistical networks with advances in automated text analysis of media content to study media exposure and Prof Lazer presents tools for analysis. |
| Year(s) Of Engagement Activity | 2016 |
| Description | Current state of the art, and future directions for linguistic analysis in a security context |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Policymakers/politicians |
| Results and Impact | Presented and discussed my work on using co-occurrence networks to analyse political text. The talk sparked a question and answer session from other academics in attendance, as well as from policy practitioners. |
| Year(s) Of Engagement Activity | 2017 |
| Description | Delivered "Text as Data" to MSc students at Exeter |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Postgraduate students |
| Results and Impact | I am currently offering a module on "Text as Data" at the postgraduate level. |
| Year(s) Of Engagement Activity | 2018 |
| Description | EXPONet Final Conference - Turing Institute |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | The presentations and debate were focused on the following objectives: Identify main challenges to providing quality information. Identify methods to enhance the quality of information and debate. Methods could lead to policy recommendations. Engage the audience in discussion |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://www.turing.ac.uk/events/measuring-information-exposure-dynamic-and-dependent-networks |
| Description | Exeter-Duke-IBM Data+ Workshop |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | In April 2017 we organized a data science workshop to foster collaboration between our team, the Duke University Information Initiative and the IBM Center for Advanced Studies in Amsterdam. Participants from each research team presented their work, and the presentations were followed by discussions with Exeter academics, postgraduate and undergraduate students. As a consequence of the workshop, we've continued collaborating with the IBM Center for Advanced Studies. |
| Year(s) Of Engagement Activity | 2017 |
| Description | GW4 Data Intensive Presentations |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Other audiences |
| Results and Impact | The team members presented their work at the GW4 Data Intensive Research Workshops in Exeter, Bristol, Bath and Cardiff and established connections with researchers from multiple other universities, who are working on similar topics. |
| Year(s) Of Engagement Activity | 2017 |
| Description | IBM CAS Amsterdam Initial Visit |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Dr. Cioroianu presented our research about measuring information exposure in online news articles at the IBM Center for Advaced Studies in Amsterdam. As a consequence, we've started working with CAS on developing a mobile news reading app and the algorithms that underpin it. Multiple working visits followed and Dr. Cioroianu received an IBM Faculty Award as well as an ESRC IAA Fellowship to continue working with IBM. |
| Year(s) Of Engagement Activity | 2017 |
| Description | IBM CAS Amsterdam University Days |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Dr. Cioroianu presented our work during the IBM Europe University days, in October 2017. The audience were IBM leaders and staff, undergraduate and postgraduate interns as well as academics from Dutch universities. with which we've started collaborating. Dr. Cioroianu has been supervising a computer science Master's student from the University of Delft during her internship with the IBM. The students worked on text sumarization , topic extraction and ideology scaling. |
| Year(s) Of Engagement Activity | 2017 |
| Description | Keynote Lecture - Institute for Policy Research |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Public/other audiences |
| Results and Impact | Gave keynote address at the Institute for Policy Research's (Bath University) 2017 symposium "Politics, Fake News and the Post-Truth Era". My address focused on the implications of social media and misinformation in campaigns. The title of the talk was "Fake News, Experts & the Failure of Political Responsibility." I reported on results of research from the EXPONet project. There were over 100 audience members and the address was followed by questions about the role of misinformation, possible policy responses and how the structure of the media system has changed with the introduction and growth of social media. |
| Year(s) Of Engagement Activity | 2017 |
| URL | http://www.bath.ac.uk/events/politics-fake-news-and-the-post-truth-era/ |
| Description | Knowledge Exchange Seminar Black Swan |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | Local |
| Primary Audience | Industry/Business |
| Results and Impact | The working paper by Williams, M. J., Cioroianu, I. & Williams, H. T. P. (2016) Different news for different views: Political news-sharing communities on social media through the UK General Election in 2015 was presented at the International Conference on Web & Social Media (Cologne, 2016) and has formed part of several seminars given by Co-Investigator Williams. Specifically the paper formed the basis of a knowledge exchange seminar at Black Swan - a big data firm focused on predicting consumer behaviour. |
| Year(s) Of Engagement Activity | 2016 |
| Description | NCRM Autumn School 2017 |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | Prof. Banducci and Dr. Cioroianu presented some of the methods and tools we've developed and used to analysed online information exposure to early career researchers and Ph.D. students attending the NCRM Autumn School in Amsterdam. The NCRM also asked us to record a version of the talk which will be used as teaching material for the NCRM website. |
| Year(s) Of Engagement Activity | 2017 |
| URL | https://www.ncrm.ac.uk/resources/video/?title=Online%20data%20sources%20linking%20old%20and%20new%20... |
| Description | NCRM Online Data Sources tutorial |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Professional Practitioners |
| Results and Impact | Prof. Banducci and Dr. Cioroianu recorded a video tutorial for the NCRM on linking clickstream, survey and social media data, and extracting and analysing information from online news articles. The video is now part of NCRM's repository of online learning resources. As a consequence of this collaboration, Dr. Cioroianu will be recording several other teaching materials for the NCRM in June 2018. |
| Year(s) Of Engagement Activity | 2017 |
| URL | https://www.ncrm.ac.uk/resources/online/online_data_sources/ |
| Description | NCRM/Exeter Computational Communication Methods Spring School |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | Regional |
| Primary Audience | Postgraduate students |
| Results and Impact | The NCRM/Exeter Computational Communication Methods Spring School will provide training at introductory and advanced levels, catering for both social scientists and data scientists. The school will take place at the University of Exeter over two 4-day sessions between 6-14 April 2022 and is co-sponsored by:- IDSAI Computational Social Science Social Data Science Group, Turing Institute Exeter Q-Step The programme will cover multiple computational approaches, such as machine learning and network analysis, and their application to communication research looking at text, images and social media data. World-leading experts will deliver workshops, seminars and demonstrations, help desks will offer one-to-one consultations and there will be opportunities for more informal networking. |
| Year(s) Of Engagement Activity | 2022 |
| URL | https://www.ncrm.ac.uk/training/show.php?article=11847 |
| Description | Panel discussant for "Voter Turnout in Portugal: diagnosis and possible solutions" a public event. |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Policymakers/politicians |
| Results and Impact | The conference was a public event that address the following issues: Huge gains in social welfare have coincided with a significant increase in abstention. Why? How? Is this a Portuguese phenomenon or part of a trend of Western democracies? Is democracy a victim of its own success? What is the meaning of not voting: alienation and indifference? Or is it a silent protest language? Why won't young people vote? What social, economic and territorial inequalities may influence electoral participation? What implications can different methods of measuring voter turnout have on our perception of electoral participation? How can the dynamics of emigration affect the official values of participation rates? The audience was made up of school students, the public, elected official and academics. A robust debate was sparked by the discussion about technology and voting. I was on a penl discussing technology and voting. |
| Year(s) Of Engagement Activity | 2018 |
| URL | https://www.pttalks.pt/en/the-conference/ |
| Description | Podcast on Different news for different views: Political news-sharing communities on social media through the UK General Election in 2015 |
| Form Of Engagement Activity | A broadcast e.g. TV/radio/film/podcast (other than news/press) |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Postgraduate students |
| Results and Impact | How are we influenced by the information we are exposed to? - Iulia Cioroianu The media environment, including the way we consume our news has been radically changed by the advent of the Internet. What does this mean for the type of content we look at and how we share it with our off- and online networks? And how does it influence our opinions? Iulia Cioroianu from the University of Exeter discusses research undertaken as part of the NCRM-funded ExpoNet project which is producing a set of tools to make it possible to examine these ideas more closely |
| Year(s) Of Engagement Activity | 2016 |
| URL | https://www.ncrm.ac.uk/resources/podcasts/view.php/media-environment |
| Description | Presentation at Labour Together research event "Political Bots & Post Truth: What is Fake New and how can we fight it?" |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Policymakers/politicians |
| Results and Impact | Address. "Political Bots & Post Truth: What is Fake News & how can we fight it?" Event Organised by Labour Together. 21 February, 2018. Cromwell Green London |
| Year(s) Of Engagement Activity | 2018 |
| Description | Program Committee for International Conference on Computational Social Science IC2S2 2018) |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Other audiences |
| Results and Impact | I am on the program committee for International Conference on Computational Social Science IC2S2 2018). |
| Year(s) Of Engagement Activity | 2018 |
| URL | https://www.kellogg.northwestern.edu/news-events/conference/ic2s2/2018.aspx |
| Description | Research featured in BBC Points West "Have We Got News for You |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Media (as a channel to the public) |
| Results and Impact | Our research was featured on BBC Points West "Have We Got News for You" 31st October, 2017 an event hosted by David Dimbleby and sponsored by Royal Television Society. Our findings and visualisation of socialbots was used by producer Anne Farrant (BBC) |
| Year(s) Of Engagement Activity | 2017 |
| URL | https://rts.org.uk/article/have-we-got-news-you-asks-bbc-points-west |
| Description | The Hashtag Election? Modeling Information Flows in the 2015 General Election |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Other audiences |
| Results and Impact | Presentation of research paper at the joint sessions of the European Consortium for Political Research Workshops, Social Media and the Dynamics of Public Opinion: New Avenues for Research, April 2016, Pisa, Italy. |
| Year(s) Of Engagement Activity | 2016 |
| Description | presentation at Social Science in the Open |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Professional Practitioners |
| Results and Impact | The aim of this workshop is to examine how social scientists and those involved in open investigations might learn from each other by establishing a dialogue focused on how combinations of techniques can add depth and richness to the sources of data we work with. This event will involve a mix of presenters from the social sciences, the arts and humanities and organisations involved in open investigations. We will think about 'opening things up' as a potent way of linking that work and explore methods for locating and analysing open data ethically but also in ways that make an impact on wider public debates. |
| Year(s) Of Engagement Activity | 2019 |
| URL | https://www.researchgate.net/publication/338855901_Social_Science_in_the_Open_An_NCRM_Innovation_Eve... |