RESEARCH ASSISTANT COMPUTER SCIENCE JOB UK (Oct-2023)
Swansea University, Bay Campus seeking Research Assistant in Computer Science (Oct-2023)
We have an exciting opportunity for those who wish to make a substantial societal impact using Machine Learning & Deep Learning technologies. We are looking for a Computer Science research assistant to join our team which includes other Computer Scientists and Psychologists on an interdisciplinary & international project. The project is funded by the Morgan Advanced Studies Institute and it is titled “Creating Accessibility and Diversity in Digital Visual Language”. You, as the research assistant, will help us develop both predictive and generative Machine Learning models based on existing data from our studies. It is 0.5 contract (17.5 hours per week) from 1 December 2023 to 31 May 2024 at Swansea pay scale Grade 7 Point 26 (£31,411 per annum) with fully covered USS pension benefits.
Faculty/Directorate:
Faculty of Medicine, Health and Life Science
School:
School of Psychology
Contract Type:
Fixed Term
Closing Date:
12-11-2023
Salary: Grade 07 £32,982 to £37,099 per annum together pro rata with USS pension benefits
Contract: This is a fixed term position for six months. Part-time 17.5 hours per week (50% FTE)
Location: Bay Campus, Swansea
Brief background to the project:
Icons are used to communicate ideas and instructions and given the remarkable rise of useful technology in our everyday lives, icons are more prevalent than ever before. Yet, our understanding about how well icons communicate ideas, especially by people from different backgrounds, age, and cultures, is not well understood. Indeed, most of the current research of icon understandability (read: Icon ability) is done using graduate students at university from the same ethnic background, which only provides limited knowledge about how well icons communicate ideas.
Our ‘Iconability’ project aims at addressing these concerns by:
- Furthering our psychological understanding about how people from different cultures and backgrounds understand the meanings of different icons.
- Using machine learning, predict how well a public demographic will understand icons based upon different icon characteristics such as:
- Concreteness and Semantic Distance
- Appeal, and Complexity
- Valence and Feeling
- And Familiarity and Order of Learning
- Generate icons for a given target audience, i.e. someone of a particular culture, or age.
Research Assistant main tasks will be:
By working with other computer scientists and psychologists in the team, this job will complete stages 2 & 3 (listed above) of this project. We’re looking for a developer with a strong machine learning background to first investigate the causal relationships between icon characteristics and how they are liked and understood by people of different cultures, ages, and sex. This will involve some statistical analysis as well as some image analysis of a large amount of icon images. Using this statistical analysis, create machine learning models to predict how well people from different demographics will understand and like icons.
The second part of the job will be to generate icons as described in stage 3. Using the latest Deep Learning research techniques, we would like to be able to generate understandable icons for a given target audience using the results of statistical analysis from stage 2 of the project. This process of generation of icons could be combined into a web application that others could sign up to and use for their projects.
Specific duties of the RA will be:
- Perform statistical analysis of existing psychological research participant data.
- Create Machine Learning predictive models based on existing psychological research participant data.
- Using existing Deep Learning models (i.e., Diffusion models) create novel icons.
- Help the team Deploy a web application to allow other developers and designers to create novel icons themselves.
The University is committed to supporting and promoting equality and diversity in all of its practices and activities. We aim to establish an inclusive environment and welcome diverse applications from the following protected characteristics: age, disability, gender reassignment, marriage and civil partnership, pregnancy and maternity, race (including color, nationality, ethnic and national origin), religion or belief, sex, sexual orientation.
Last Date:
November 12, 2023
