Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Investigating the Reciprocal Relationship between News and Parliament: A Study of Strikes in the UK Using Deep-learning Sentiment Analysis and Vector Autoregression

Meijers, Yente LU (2023) SIMZ51 20231
Graduate School
Abstract
Previous research has shown that the relationship between the news and parliament is complex and highly variable. Studies have found that in many countries, parliament has minimal influence on the news. However, in the UK, a reciprocal influence between parliament and the news was found, with the media having a stronger effect. This thesis investigates how the news and parliament affect each other, using the recent strikes in the UK as a case study. The relationship is established through both the frequency of newspaper articles and parliamentary speeches and the sentiment conveyed in these texts. Additionally, the political alignment of Members of Parliament and newspapers is considered in this relationship. This thesis introduces an... (More)
Previous research has shown that the relationship between the news and parliament is complex and highly variable. Studies have found that in many countries, parliament has minimal influence on the news. However, in the UK, a reciprocal influence between parliament and the news was found, with the media having a stronger effect. This thesis investigates how the news and parliament affect each other, using the recent strikes in the UK as a case study. The relationship is established through both the frequency of newspaper articles and parliamentary speeches and the sentiment conveyed in these texts. Additionally, the political alignment of Members of Parliament and newspapers is considered in this relationship. This thesis introduces an innovative methodology by using an open-source deep-learning Natural Language Processing model (SiEBERT) for analysing the sentiment in the two text sources, combined with vector autoregression. Vector autoregression models estimate the relationship between the number of articles and speeches and the influence of sentiment and political alignment.
The results indicate that the relationship between the news and parliament is indeed reciprocal, and both have a similar effect size. Importantly, the news reacts much more quickly to parliament than vice versa. Also, the thesis finds that negative news articles and speeches are much more influential than positive ones. The political party of the speaker and the political leaning of the newspaper play an important role in the relationship, with the effects varying across ideologies. This study also concludes that both the majority of news articles and parliamentary speeches are very negative towards the strikes in the UK. (Less)
Please use this url to cite or link to this publication:
author
Meijers, Yente LU
supervisor
organization
course
SIMZ51 20231
year
type
H2 - Master's Degree (Two Years)
subject
keywords
parliamentary debate, Hansard, newspaper coverage, UK strikes, agenda setting, BERT, SiEBERT, NLP, deep learning, sentiment analysis, vector autoregression
language
English
id
9126213
date added to LUP
2023-06-21 14:12:19
date last changed
2023-06-21 14:12:19
@misc{9126213,
  abstract     = {{Previous research has shown that the relationship between the news and parliament is complex and highly variable. Studies have found that in many countries, parliament has minimal influence on the news. However, in the UK, a reciprocal influence between parliament and the news was found, with the media having a stronger effect. This thesis investigates how the news and parliament affect each other, using the recent strikes in the UK as a case study. The relationship is established through both the frequency of newspaper articles and parliamentary speeches and the sentiment conveyed in these texts. Additionally, the political alignment of Members of Parliament and newspapers is considered in this relationship. This thesis introduces an innovative methodology by using an open-source deep-learning Natural Language Processing model (SiEBERT) for analysing the sentiment in the two text sources, combined with vector autoregression. Vector autoregression models estimate the relationship between the number of articles and speeches and the influence of sentiment and political alignment.
The results indicate that the relationship between the news and parliament is indeed reciprocal, and both have a similar effect size. Importantly, the news reacts much more quickly to parliament than vice versa. Also, the thesis finds that negative news articles and speeches are much more influential than positive ones. The political party of the speaker and the political leaning of the newspaper play an important role in the relationship, with the effects varying across ideologies. This study also concludes that both the majority of news articles and parliamentary speeches are very negative towards the strikes in the UK.}},
  author       = {{Meijers, Yente}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Investigating the Reciprocal Relationship between News and Parliament: A Study of Strikes in the UK Using Deep-learning Sentiment Analysis and Vector Autoregression}},
  year         = {{2023}},
}