This inter-disciplinary project investigates the use of algorithms for coding, labelling and extracting information from the images used by political actors and interest groups in online political communication.
A third of all social media posts include images, and the posts related to government and politics which contain an image generate the highest levels of public engagement. Research in psychology shows that visual stimuli have a stronger impact on individual perceptions than other types of stimuli. In addition, the political communication literature shows that visuals can contain more information than other types of content, and that political actors and interest groups can use images to convey information that they would not otherwise convey in writing or verbally.
Recent developments in the field of computer vision and new techniques for automated visual content analysis made it easier to acquire, process and extract meaning from images, and these methods are increasingly being used in academic research, but also in industry, by online and social media companies and platforms. Social scientists have been focusing on the content analysis of political images, the way they are being used by different political actors, and the effects they may have on public opinion. Online and social media platforms and companies on the other hand have been more interested in ensuring that the images uploaded by users abide by the platforms’ rules and terms of service and are not in contravention of existing legislation (for example by promoting discrimination or extremist political views).
The proposed project builds upon and advances these two lines of enquiry by addressing the following research questions:
(1) What algorithms, methods and tools can be used to analyse visual content?
(2) How have these methods been adapted to analyse political and ideological visual content? What are the limits and implications of specific forms of automation applied to this context?
(3) What are the social, political and technological factors shaping online platforms’ decisions to remove certain types of visual content? What are the factors shaping their choice of algorithms and selection of features, and how do their automation choices ultimately lead to the decision to keep or remove certain types of visual content?
(4) What strategic incentives do platform design, automation and content restriction policies provide for political actors and interest groups and how are they modifying their communication strategies in response to these constraints?
(5) How is accountability, authority and responsibility for decision-making distributed in these contexts, and what is the role played by the legal and public policy context?
This project is associated with the UKRI CDT in Accountable, Responsible and Transparent AI (ART-AI), which is looking for its first cohort of at least 10 students to start in September 2019. Students will be fully funded for 4 years (stipend, UK/EU tuition fees and research support budget). Further details can be found at: http://www.bath.ac.uk/research-centres/ukri-centre-for-doctoral-training-in-accountable-responsible-and-transparent-ai
The project will contribute to explanations of how AI affects the social world and it will use social scientific concepts to do so. Its empirical discussion will also contribute to computer science and engineering understandings of the ethical, political and social implications of machine-learning and/or automotive design in politics and political communication.
Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or equivalent). A master’s level qualification would also be advantageous.
In addition, it is anticipated that applicants will have a background in a social science related subject, such as sociology, politics and/or geography. They should have a ‘B’ in A-level Maths (or equivalent) to enable them to take full advantage of the interdisciplinary training. Prior knowledge and work on automated decision-making would be an advantage but is not essential. Students will receive training tailored to their background and project. This may include: programming, digital data and advanced quantitative methods (social statistics), AI ethics and risk, AI, politics and government.
Informal enquiries about the project should be directed to Dr Iulia Cioroianu on email address [email protected]
Enquiries about the application process should be sent to [email protected]
Formal applications should be made via the University of Bath’s online application form for a PhD in Computer Science: https://samis.bath.ac.uk/urd/sits.urd/run/siw_ipp_lgn.login?process=siw_ipp_app&code1=RDUCM-FP01&code2=0013
Start date: 23 September 2019.