Sentiment Analysis of LegCo Members' Speech on Chief Executive

Sentiment analysis on lawmakers' speeches in Chief Executive-attended Hong Kong Legislative Council meetings

What it does

A sentiment analysis on the words used by lawmakers in LegCo meetings that the Chief Executive attendeds. The analysis examines data from the Fifth LegCo (2012-2016) and the Sixth LegCo (2016-2019) up until May 2019, which stretches across the CE terms of Leung Chun-ying and Carrie Lam.

What they are

• fetch_data.R: Scripts to fetch a lawmaker list and the speech text from the LegCo API and and export as cy_legco_text.rds, cl_legco_text.rds and members_name.rds
• preprocessing.R: Functions to compute the sentiment of the text
• sentiment_member.R: Scripts to compute sentiment of each lawmaker's speech and plot bar graphs in English and Chinese as cy_en.png, cy_zh.png, cl_en.png and cl_zh.png
• sentiment_timetrend.R: Scripts to compute sentiment the speech the overall sentiment of each meeting and plot time trend in English and Chinese as timetrend_en.png and timetrend_zh.png.
• pro_est5.rds & pro_est6.rds: Lists of pro-establishment lawmakers in the Fifth and Sixth LegCo

How it works

The Chief Executive attends LegCo during CE's Question and Answer Sessions on the Policy Address, CE's Question and Answer Sessions and CE's Question Time. When putting questions to the CE, lawmakers often take this opportunity to express their views about the administration and/or the CE.

This analysis matches sentiment terms from the Bing lexicon (for English text) and the National Taiwan University Semantic Dictionary (for Chinese text) with the speech text and caluclate the proportion of positive and negative terms used. This can provide a glimpse of the relationship between the legislature and the executive body.

What you need

You need the R packages elgarteo/legco and elgarteo/legcoplus to fetch data from the LegCo API.

All other packages are obtainable from CRAN.