Clustering emotions of people in various Indian cities using Instagram

Abstract

In the growing age of social media, people express their emotions in the form of pictures, videos and text. And these emotions often affect a person’s life in many aspects. The main aim of this study was to examine how people of the most literate cities of India post different emotions through pictures on Instagram. In our work we have created a dataset of 60,000 Instagram posts for the top 200 literate cities in India, 300 posts for each city. The analysis of this dataset shows that 80% of the images consist of happy emotion, neutral in 82%, sadness in 43%, contempt in 47%, surprise in 20%, and for disgust and fear it is 9% and 6% respectively.

Subject Classification:

• (2010) 68T05

Keywords:

• Emotion analysis
• Human emotions
• Facial recognition
• Social media
• Instagram
• Azure
• Data analysis
• Machine learning
• Data science