Posted: 24/02/16

With just two days left to polling day and all of the opinion polls showing unclear outcomes, research conducted by the ADAPT Centre for Digital Content Technology provides insights into public opinion on Twitter and helps give an indication of what way the election may sway.

As the four main party leaders clashed on RTÉ news, a system developed by ADAPT researchers tracked Twitter’s reaction.  For each of the leaders the number of tweets sent in support (positive) and in criticism (negative) were counted, as well as which issues were being discussed.  

Taking part were: Taoiseach and Fine Gael leader Enda Kenny; Tánaiste and Labour leader Joan Burton; Fianna Fáil leader Micheál Martin; and Sinn Féin leader Gerry Adams.

As we move towards polling day when Ireland will decide who will govern Ireland for the next five years, the ADAPT Centre’s analysis provides a view on how citizens are reacting to the leaders, their policies and the key issues.

Who won according to the Twitterati?

The key insight was the reduction in reaction on Twitter to the final TV debate with significantly reduced mentions for all leaders.  Mention also needs to be given to the floor ‘creak’ which seems to have taken Twitter by storm!  However as a non-official election candidate reaction was not scientifically analysed by ADAPT!

Looking first at individual leaders, conversations about Gerry Adams dominated the discussion with almost twice the number of tweets to the next most tweeted about leader.  Sentiment in relation to the Sinn Fein leader was almost even, with 37% positive and 36% negative, with the remainder neutral.

Both Taoiseach Enda Kenny and Tánaiste Joan Burton received in excess of 2000 tweets each during the debate.  According to Twitter, Joan’s performance was  very favourable to the Twitter population with 43% of comments classed as positive and only 29% negative.  The Taoiseach also came out more positively in the tweets but that was only a marginal lead with 41% of tweets positive and 40% negative and the remainder neutral.

Of all the leaders, Micheál Martin received the fewest amounts of comments on Twitter with just 1,131 tweets.  Sentiment was mostly favourable however with  41% of those classed as positive and  and 34% negative with the remainder neutral.

Biggest Topics

Unsurprisingly, the top topics were finance, health and housing. Interestingly 47% of the conversations on healthcare were positive with 35% negative and the remainder neutral.  Housing was the next biggest cause for reaction on Twitter with 55% of comments negative and only 24% positive.

The system that has been developed within ADAPT can quickly track the volume of tweets and gauge reaction and sentiment from the tweets in relation to election topics, parties and candidates. The project presents a graphical analysis of tweets that can be filtered by party or candidate allowing the user quickly understand and gain insights into the conversations on election topics taking place on Twitter.

The research team developed a system that harvests and filters tweets using key hashtags, the main party hashtags, candidate twitter handles, names, and expected election topics along with associated keywords and synonyms. This information is then presented in a series of graphs highlighting top issues, candidate mentions, party mentions, and sentiment to government and opposition parties. Tweets can be filtered by date, party and candidate in any combination.

The lead PI on the project was Professor Séamus Lawless from the ADAPT Centre at Trinity College Dublin’s School of Computer Science and Statistics. Professor Lawless led a team consisting of Motasim Alsayed, Haithem Afli, and Matthew Nicholson.

Share this article: