Real-time Bitcoin price tendency awareness via social media content tracking

Abstract

Cryptocurrency has been introduced as a relatively new financial system that is widely spread among traders and investors all over the globe. This type of digital currency is hugely attracting social media attention; Social media community along with investors and traders interact to share knowledge or to predict the price tendency of the market. Bitcoin, the leading cryptocurrency nowadays, has the highest market capitalization among other currencies. Which means that any major change in its price tendency will definitely affect the whole market and therefore other coins’ prices will surely rise or fall accordingly. We can assume that the greatest concern ever of all the traders around the world is to be alerted or aware of such major price tendency shifts in real time, in which may help them gain more profit or cut losses before it is late, or in either way predict the potential market movement of other digital currencies. It is reported that emotional interactions of the social media users especially on Twitter (one of the most globally used micro-blogging social networks especially related to cryptocurrency topics) have a great influence on the trend of the Bitcoin price. The huge number of daily active Twitter users with the enormous volume of tweets related to Bitcoin price tendency makes it remarkable regarding its impact on Bitcoin market interaction. In this paper, we will implement Apache Spark logistic regression model to process the large-scale data after having a sentimental analysis of Twitter tweets regarding Bitcoin, to predict the upcoming price tendency and classify an awareness level to alert potential traders and investors in real time about such potential market changes.