Predicting Eurovision Song Contest Results by Interpreting the Tweets of Eurovision Fans

被引:0
|
作者
Demergis, Dimitri [1 ]
机构
[1] Rowan Univ, Coll Sci & Math, Glassboro, NJ 08028 USA
关键词
natural language processing; sentiment analysis; lexicon generation; language parsing; social media; web text analysis;
D O I
10.1109/snams.2019.8931875
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The Eurovision Song Contest is Europe's most well-known international competition for musical performance and songwriting. The winner of the contest is determined by a voting system that takes into account the opinions of viewers across the continent and around the world. This paper attempts to identify a link between the overall sentiment of Twitter users toward the performances and the eventual winner selected by Eurovision voters. More than 900,000 English and Spanish-language tweets that were posted during the 2019 Eurovision Song Contest were collected for this study. The tweets were analyzed to identify the sentiment and target performance for each. The results of this analysis were used to determine a predicted rank for each performance, resulting in a Spearman rho correlation coefficient of 0.559 and a Kendall tau correlation coefficient of 0.403.
引用
收藏
页码:521 / 528
页数:8
相关论文
共 50 条