Keelan Fadden-Hopper


ENO Opera Finder Quizbot (code)

I worked as the technical lead with a team of other students to produce a prototype quizbot for the English National Opera to recommend operas that young people might be interested in watching, based on their Facebook likes and their responses to a series of questsions about their interests. The project was built using the Python web framework Django and the Facebook API.

The Electionary (site) (code)

I analysed the use of language in American presidential election campaigns from 1960 to the present day, with a group of other students on my Quantitative Methods course at UCL. I used Python and scrapy to collect and process the data from transcripts of presidential election debates, as well as using the Twitter API to collect data from the 2016 candidates' Twitter activity. We then analysed this data using sentiment analysis techniques. Finally, we produced visualisations (mainly using matplotlib), including a visualisation of geospatial data, and produced a website.

Walking to university (report)

I examined my behaviour walking to and from UCL to see which parts of my journey were faster and slower. I collected location and time data from my phone using Moves. I analysed this using Python, and visualised the results using Mapbox.

UCLSocBot (code)

A simple Twitter bot I worked on with a team at a UCL hackathon. Using the UCL and Twitter APIs, it tweets every time a university society makes a new booking for a room.