We present a dataset created from the Hansard House of Commons archived debates of the UK parliament (2013-2016). The resource includes fine-grained topic annotations at the document level and is enri