SNS Fatigue Extraction by Analyzing Twitter Data

Abstract

SNS fatigue has become a problem in SNS, i.e., Twitter and Facebook. In this work, we define it as “physical and mental fatigue caused by using SNS” This is one of the most widely used stress experiences on Twitter among young people. Therefore, we analyze the causes of SNS fatigue on Twitter to extract SNS fatigue using tweet data, we aim to create an index to determine SNS fatigue to reduce SNS stress in the future. In this paper, we collected questionnaires about how much stress was felt by 10 Twitter users on 25 events that could cause stress in Twitter usage. Then, we classified the causes of SNS fatigue into three main labels by a principal component analysis. For extracting SNS fatigue, we collect tweets and label those collected tweets to extract feature words of the tweets for each label. Also, we create a classifier for the causes of SNS fatigue using a machine learning algorithm. In this way, SNS fatigue prediction and SNS stress reduction can be expected using feature words for SNS fatigue. Finally, we verified the effectiveness of feature word extraction for SNS fatigue and the classification accuracy of the causes of SNS fatigue.