Project: Social Behaviour Analysis: Quantifying User Opinions

Project: Measurement of Opinion based on Social Networking Data

This research project explores methods for quantitative measurement of individual's opinion based on social networking data (in contrast to previous NLP methods based on text content). Different methodologies will be examined and compared, including methods from statistical analysis (Bayesian inference), network science (Correspondence Analysis) and machine learning (Neural Embedding). In particular, we aim to improve and evaluate the Neural Embedding method which is perhaps the most promising method.   

Requirements: 

• BSc/MEng/MSc student in the Department of Computer Science.

• Good knowledge and skills in statistics, mathematics, data science, machine learning and network science.

• Good skills in programming, such as Python.

More information:

• BSc-CS final year project report (2024) Title: Opinion Measurement Based on Social Connectivity Data 

The following talk slides provided a brief introduction to the three opinion measurement methods. 

• Talk slides 

The following paper introduced the Neural Embedding method for opinion measurement based on networking data. 

• "Quantifying social organization and political polarization in online platforms" Nature, Dec 2021. 

  • https://arxiv.org/abs/2010.00590   

Contact:

Please contact Shi Zhou <s.zhou (at) ucl.ac.uk> for more information.