Detecting Irony on Greek Political Tweets

The Modern Greek Language on the Social Web: A Survey of Data Sets and Mining Applications

“…The developments that have taken place in Greek political life have steered the interest of research community on recognizing political opinion while assessing the idiosyncrasy of the Modern Greek language. This work acknowledged several studies that have attempted to capture the impact of political and social developments over the last decade [8,14,[39][40][41][42][43][44]. Regarding extracting political sentiment from textual data, researchers focused on analysis before and after the events (elections, referendum, etc.)…”

“…Recent research work has indicated that Greek social media presents a platform for users to express their opinion related to many aspects of private and social life and their experience with services and products. This section presents recent literature on the political footprint along with voting patterns (Section 3.1 [8,14,[39][40][41][42][43][44]) and introduces to the reader work related to Marketing and Business Analysis (Section 3.2 [3,5]) that employ state-of-the-art opinion-mining ML techniques.…”

“…Recently, a study from Charalampakis et al [40,41] attempted to further categorize semi-supervised results from an updated version of their previous research. On the same data set 24 , they implemented the collective classification technique, a semi-supervised learning method, which allows humorous political tweets to predict actual election results.…”

“…Regarding politics and voting analysis, public opinion has been in the centerpiece of recent research works [8,[39][40][41][42][43][44]48]. We have found a plethora of related studies attempting to tackle the traditional analysis of the political domain and social behavior.…”

Analysis of evolution, dynamics and vulnerabilities of online social networks

“…Regarding noise, a huge proportion of election-related Twitter posts are humorous, ironic or sarcastic and do not portray any party (or ideology) inclination. It is estimated that approximately half of collected tweets belong to this category [22,72]. Filtering out these posts or users is a challenging task and relies heavily on qualitative human-crafted datasets of sentiment vocabularies and pre-classified, "ground truth" samples [152].…”