Analysing the Irish 2011 General Election Party Manifestos and Irish Political Statements from the Irish Parliament (D´ ail) using Text Analytics John Ryan Department of Informatics Institute of Technology Blanchardstown Dublin, Ireland Email: leabharnua@gmail.com Markus Hofmann Department of Informatics Institute of Technology Blanchardstown Dublin, Ireland Email: markus.hofmann@itb.ie Abstract—This research paper uses text visualisation methods, such as Wordclouds, to emphasis the key terms used by different political parties throughout the 2011 Irish Parliament (D´ ail) period; this includes the final sessions of the 30th D´ ail before the 2011 General Election and the entire 2011 sessions of the 31st D´ ail. Using term frequencies this research clearly shows the topics that were important to each relevant political party. This study also highlights the key differences that may reside between what a political party stated clearly in their election manifesto and what was written on D ´ ail transcript records during the course of 2011. The main objective of this research is exploring and un- derstanding ways in which key political information can be summarised and subsequently displayed in an intuitive visual fashion as a guiding education facility. I. I NTRODUCTION This research paper avails of text visualisation imagery (such as Wordclouds) to fully explore and understand the similarities and the differences between a political party manifesto and what was set on record in the Irish Parliament (D´ ail ´ Eireann). It also endeavours to analyse elements of political manifestos such as the Parts-of-Speech (POS). The data for this corpus was sourced from 2 distinct areas; the political statements of record were gathered from the Oireachtas (which would be Ireland’s National Parliament) transcripts; in particular, the 2011 transcripts from D´ ail ´ Eireann (which is the Irish equivalent to the United States House of Representatives or the United Kingdom’s House of Commons). The data transcripts for this research were obtained from the following Oireachtas online resource: http://debates.oireachtas.ie/XML/D/2011/ The political transcript data was downloaded via web crawl operators, extracted using both regular region and regular expression functions, and then stored within a mySQL struc- tured database environment. In addition to this, information relating to specific political parties was also crawled and stored in order to link each politician’s statement within the Oireachtas transcripts to the appropriate political party. The corpus was then mined for particular topics using Standard Query Language (SQL). The 2011 General Election manifestos were sourced from: • http://www.labour.ie/download/pdf/labour\ election\ manifesto\ 2011.pdf • http://michaelpidgeon.com/manifestos/byelection.htm It should be noted that Fine Gael’s 2011 Manifesto was originally available on http://www.finegael2011.com but this site is now discontinued. Also, Fianna F´ ail’s mani- festo was originally sourced from the Irish Times elec- tion website http://www.irishtimes.com/indepth/election2011/ manifesto/fiannafailManifesto.pdf but again this link is also discontinued. Currently, a consistent repository of Irish party political manifestos is http://michaelpidgeon.com/manifestos/ byelection.htm. The main analysis method that was used to achieve this study’s objectives was to analyse documents for their most often used words through mechanisms such as term frequency measurements. Thus, the most popular used words in a spec- ified window of examination can be identified. Lists of the most frequently used words (topics) can then be screened via unmatched queries, for example, in order to subtract topics that both political parties may share to find what issues are unique to a particular political party. Throughout this research paper, the political party names are interchanged by their abbreviation for expediency purposes; FF (Fianna F´ ail) and FG (Fine Gael.) The analytic software used in this research to extract the data for the political corpus was RapidMiner 5.3.005; this is a free to download application that provides text and data mining functionality and is sourced from http://rapid-i.com/content/ view/181/190/. II. METHODOLOGY A. Brief Background to the Structure of Irish Political System D´ ail ´ Eireann, Ireland’s National Parliament, is one of the two houses within the Oireachtas. The Oireachtas is structured with a President and two Houses: D´ ail ´ Eireann as previously mentioned and Seanad ´ Eireann (although it resembles the US Senate, it differs in many strategic ways, namely, the Senators 658 978-1-4799-2572-8/14/$31.00 c 2014 IEEE