Social networks such as Twitter contain billions of data of users, and in every second, a large number of tweets trade through Twitter. Sentiment analysis is the way toward deciding the emotional tone behind a series of words that users utilize to understand the attitudes, thoughts, and emotions that are enunciated in online references on Twitter. This chapter aims to determine the user preference of Bitcoin and Ethereum, which are the two most popular cryptocurrencies in the world by using the Twitter sentiment analysis. It proposes a powerful and fundamental approach to identify emotions on Twitter by considering the tweets of these two distinctive cryptocurrencies. One hundred twenty thousand (120,000) tweets were extracted separately from Twitter for each keyword Bitcoin/BTC and Bitcoin/ETC between the period from 12/09/2018 to 22/09/2018 (10 days).
TopIntroduction
The footstep in big data analysis is to accumulate the data. This is called “data mining” (Hewage et al., 2018). Those data can be from any source. There are lots of data sources where we can collect a huge number of data. Twitter is one of the best sources used in data science. It is also free social networking media that allows users to broadcast tweets. These tweets are known as short messages used to send tweets (short messages) for all kind of reasons such as pride, attention, dullness, help, to become famous etc. The vast majority of users use Twitter for fun, giving a shout out to the world and ensure to distribute ideas within communities. Unlike other social platforms, every user's tweets are entirely public and able to retrieve. Twitter data can be accessed by general public perception and how they sense topics. Twitter's API allows developers to pull data. However, some limitations are included (Wu et al., 2011).
Sentiment analysis is the process of determining the emotional tone behind a series of words that can be used to understand the attitudes, thoughts, and emotions articulated in online references. It is very beneficial in monitoring social media (Singh et al., 2018; Hewage et al., 2018) since it enables us to present broader public opinion about specific topics. It can also be a significant part of market research and approach to customer service.
Cryptocurrency is a digital or virtual currency designed to work as a medium of exchange that uses cryptography for security. Cryptocurrency is also an equivalent electronic currency (Bohr et al., 2018). Bitcoin has a rapid rise in the price of a virtual currency over the past few months. The basis for creating Bitcoin and all subsequent virtual treaties is to address some perceived deficiencies by way of payment being made from one party to another. The most famous cryptocurrency is bitcoin which makes everyone interested in the matter of encryption, because of its unpredictable growth that has become the de facto standard for cryptocurrencies (Kaushal et al., 2017). Although there are various cryptocurrencies, Bitcoin and Ethereum are the most two that lead the market. Bitcoin (BTC) is the first coin, and Ethereum (ETH) becomes the second a few years later. However, Bitcoin and Ethereum have different purposes. Bitcoin is created as a substitute for regular currency; it is a medium for payment transactions and value storage, while Ethereum is developed as a platform to promote peer-to-peer contracts and applications through its currency instruments (Velankar et al., 2018).
Most of the reviewed articles use Twitter API to retrieve data from Twitter while some articles used paid APIs such as Sentiment 140 (Chakraborty et al., 2017; Attarwala et al., 2017). Some articles find out that there is a high correlation between the probabilities of the influence of Twitter users and the probabilities influenced, and most users maintain a sentimental balance in both. Twitter provides the functionality of social networks is used to know if users are exposed to the surroundings in the online social world that control their sentiments. The designed models to learn both sentimental influencing probabilities and influenced probabilities for users and present observations are based on real social network data. For each posted tweet by a user, an emotional analysis is performed to determine its polarity, whether it is positive or negative. After pre-processing the raw data, followed classifiers NB, SVM, N-SVM, ME, KNN, DT are used. Information Gain and Gain Ratio techniques are used for feature reduction. Knowledge enhancer and synonym binder module are applied to enhance the information again. This chapter discusses how to use Twitter data to choose the user preference between Bitcoin and Ethereum, which is useful for people who are new to cryptocurrencies to get ideas before they get involved.
The rest of the chapter is organized, as follows: Section 2 introduces the dataset and materials and methods used. Subsequently, the Tweet comparison results on Ethereum and Bitcoin are presented in Section 3. Section 4 provides a related discussion and potential future improvements in the area, and, finally, the chapter concludes in Section 5.
TopMaterial And Methods
Figure 2. Procedure of the research