Sentiment Detection in Scotland: MU at ACM Conference
The ACM Conference on Information and Knowledge Management is an internationally renowned conference to present the latest developments and innovations in information retrieval, knowledge management, and databases. Since its foundation 20 years ago, it has been a key annual meeting of the minds for leading scientists in this field.
Stefan Gindl, researcher and lecturer at the Department of New Media Technology, had the chance to join the over 700 participants this year in Glasgow, Scotland. He presented cutting-edge research in the area of sentiment detection, which has the challenging goal of making computers interpret human opinions by analyzing their writings. This research has been incorporated as an integral component into award-winning public showcases such as the"Media Watch on Climate Change".
Sentiment detection methods require a list of opinionated terms, a "sentiment lexicon". In order to collect a solid database for analysis, MU's new media team harnessed the power of social media and developed a game-with-a-purpose to create a sentiment lexicon from scratch. This Facebook application called the "Sentiment Quiz", presents various terms to human players and asks them for their opinion whether the term expresses positive or negative sentiment. The quiz keeps the players engaged by rating terms such as "wail" and "smile" for positive and negative connotations, which improves the coverage and accuracy of the lexicon.
To the backdrop of craggy Scottish hills, the stimulating atmosphere of the conference allowed for an exchange of ideas among scientists of many different nations. The social events offered by the organizers supported the deepening of fresh acquaintances in a recreational surrounding after a long day of scientific inspiration. Back home at MU, Stefan has returned with fresh ideas to further improve the new media department’s efforts on sentiment detection.
[img_assist|nid=3847|title=|desc=|link=none|align=center|width=457|height=500]
