Sentiment correlation discovery from social media to share market

Social media data analytics have been successfully applied in many real applications such as product recommendation, target advertisement. In recent years, it also attracted lots of attention from the financial researchers to analyse the financial trending or stock marketing prediction. In this paper, our goal is to investigate the meaningful way of uncovering the correlation between the stock share price change and the social media data usage. In this work, we first provide a mechanism to collect Twitter data, use Latent Dirichlet Allocation for topic modelling, then perform the sentiment analysis based on topics, and finally discover the correlation between social media and share price. Based on our empirical results, we find that the correlation could be impacted by the popularity of discussion as well as the valence of community, which represents the happiness to the target companies to be analysed and predicted. This could be built up by exploring the market and crisis resolution. The influence of online social users also plays a significant role in the correlation, which is a factor of manipulation that the influential users should be considered by measuring the responsibility of their social media account.