Taught Courses
This year (2011-12) I am teaching XML & Structured Information (AMCM043), Interaction Design (DCS318/AMCM318), Interactive Systems Design (AMCM016) and the MEng Team Project (ELE614). See here for my office surgery hours.
Materials for the courses are available here:
Postgraduate Projects
Some possible ideas for MSc or PhD projects:
- Measuring festival audience engagement via social media
- Building and deploying a festival audience engagement mobile app
- Investigating distributional approaches to modelling dialogue
- Improving information extraction for dialogue data
- Theoretical modelling of how people talk and interact on Twitter
- Automatic understanding of language and interaction on Twitter
- Automatic detection of some interesting dialogue phenomenon (e.g. "split" utterances, or clarification requests)
- A comparative study of clarification behaviour in Chinese (or a similar language) and English
- Investigating incremental planning for incremental natural language generation (PhD only)
- Improving our incremental dialogue system DYLAN (PhD only)
- Investigating grammar induction for Dynamic Syntax (PhD only)
Undergraduate Projects
Some possible ideas for BSc projects:
- Building a question-answering system for visitors to the EECS website
- Building a spoken or text-based dialogue system to give information about EECS
- Investigating distributional/probabilistic models of how people speak in dialogue
- Investigating statistical models of how people talk on Twitter
- Designing and building an automatic chatbot for Twitter
- Designing and building an automatic question-answering system for Twitter
- Improving our automatic sentiment/emotion detection algorithms
- Experimenting with text analysis to improve multimedia data retrieval (e.g. can we use Youtube comments to improve image retrieval?)
- Graphical visualisation of dynamic social media (e.g. Twitter) data around a topic (e.g. can we tell what people thought of the various acts at Glastonbury based on what they tweeted?)
- Building a mobile phone app which uses our sentiment/emotion detection algorithms to display something interesting about a user's data, location, Facebook status or similar