Data Science at LIDA: Social Sciences

The 1st instalment of our data science mini-series showcasing current research projects here at Leeds Institute for Data Analytics (LIDA). This weeks features 10 minute lightning talks from three data scientists about their research on Social Sciences.

Can We Predict Loneliness?
Aditi Sudhakar

In the wake of the Coronavirus pandemic, loneliness has emerged as a pressing societal concern, especially with strained mental health support services. This project, in collaboration with the Bradford Institute for Health Research, seeks to understand loneliness using anonymised patient data and self-reports of loneliness using data from the Understanding America Study (UAS). By investigating correlations with physical and psychological symptoms, the study aims to create visualisations predicting loneliness-prone areas based on socio-economic factors and lay the groundwork for future agent-based modelling studies to examine the spread of loneliness across society. The findings will aid policymakers in crafting effective strategies and provide individuals with tailored coping mechanisms to improve resilience and well-being in a post-pandemic world.

Socio-Economic Patterns of Household Waste Generation
Favour Aghaebe

Our project, using West London household waste data analyses patterns in household avoidable food waste generation and seeks to relate this to demographic and social causes. First, we create a working dataset which comprises information relating to average household waste as well as demographic information of an area. This is achieved by combining output area level collected data with publicly available census data. Then using a combination of statistical techniques and methods, we identify statistically significant demographic variables that contribute to the increase or decrease in avoidable food waste generation. By exploring the socio-economic patterns of household waste generation, this project aims to help inform initiatives and targeted policy interventions to reduce waste, emissions and pollution as well as contribute towards the realisation of SDG 12, responsible consumption and production.

Developing eco-labelling for food sold on campus
Lydia Wharton

In partnership with the Sustainability Service and working with the catering department, this data science project focuses on developing eco-labels for food sold on the University of Leeds campus, promoting environmentally sustainable choices. Conventional eco-labelling often reports environmental factors as single values, disguising large variabilities due to production location and techniques. This research assesses the environmental impact of food consumed at the university using Monte Carlo simulations to analyse variability in emissions, land, and water use of campus recipes, leading to more accurate eco-labels. A Flask-based app is being developed to manage recipe data, automate environmental impact calculations, and print labels, supporting the University's goal of becoming net-zero by 2030.

Host or Publisher Leeds Institute for Data Analytics