Mining the Relationship between UK Parliament and European Parliament Debate Data

Chaohai Ding

During the second Talk of Europe creative camp, we have enriched the Talk of Europe data in three ways:

  1. 5-star Linked Data: we used named entity recognition algorithm to automatically match the same MP in the Talk of Europe dataset, WhatTheySaid dataset, and DBpedia dataset. Then we used the owl:sameAs assertion to interlink the same entities.
  2. Mining the background information: after interlinking the same entities, we used data mining algorithms to cluster the MPs based on the topics of debate, sentiments and their background information (such as education, university attended, major subjects)
  3. The relationship between UK parliament and European Parliament: This is also what we wanted to do in the first creative camp. This time, we looked into the relationship of the UK and EU parliament debate after we interlinked the two datasets together. We mainly used the topic extraction, summarization algorithms and sentimental analysis to find the similarities of debates between two datasets.

Click here for a presentation of all the results.
Click here for the online demo.