Automatic extraction of relevant keyphrases for the study of issue competition

EasyChair Preprint 875, version 2

Abstract

The use of keyphrases is a common oratory technique in political discourse, and politicians often guide their statements by recurrently making use of keyphrases. We propose a statistical method for extracting keyphrases at document level, combining simple heuristic rules. We show that our approach can compete with state-of-the-art systems. The method is particularly useful for the study of policy preferences and issue competition, which relies primarily on the analysis of political statements contained in party manifestos and speeches. As a case study, we show an analysis of Portuguese parliamentary debates. We extract the most used keyphrases from each parliamentary group speech collection to detect political issue emphasis. We additionally show how keyphrase clouds can be used as visualization aids to summarize the main addressed issues

Keyphrases: Information Retrieval, keyphrase extraction, political discourse

@booklet{EasyChair:875,
  author    = {Miguel Won and Bruno Martins and Filipa Raimundo},
  title     = {Automatic extraction of relevant keyphrases for the study of issue competition},
  doi       = {10.29007/mmk4},
  howpublished = {EasyChair Preprint 875},
  year      = {EasyChair, 2019}}