Abstract
Clubhouse is an invitation-only social networking application that differs from the usual social media platforms in that it is “audio only.” In this chapter, the sentiments in the social media messages about Clubhouse in the classic SMPs are examined by supervised learning (by using Hugging Face Transformer Library), and the user feelings are analyzed. Because Turkey is in the first ranks among European countries in terms of both the number of social media users and the number of messages, the analysis is conducted using the Turkish users. Mentions of Clubhouse have begun on Twitter and Sourtimes platforms in Turkey in early 2021. In this study, the aim is to demonstrate how Clubhouse, a new and different SMP, is evaluated by Twitter and Sourtimes users and to reveal user thoughts about this SMP along the timeline by using sentiment analysis.
TopIntroduction
Clubhouse, which was launched on the App Store on 17.03.2020, differs from the usual social media platforms in that it is based on “audio”. Although the application was only available for the iOS operating system until May 2021 and participation is on the basis of invitation, which caused a limited number of participants for a long time, the Lion King event, which was held on 27.12.2020 and where the casting was done on the Clubhouse platform, spread by the word of mouth and increased the number of users. At the beginning of 2021, in Turkey, users started to talk about Clubhouse on Twitter and Sourtimes platforms, especially the Turkish users living in the USA. Since the second half of January 2021, people started to see the Clubhouse invitations, which aroused great curiosity and attracted users, especially with the viral effect of Twitter. Clubhouse invitations were also offered for sale on e-commerce sites such as “Ebay”. The platform, which increased the number of participants in Turkey as well as in the world in a very short time with the participation of celebrities, has active participants from many different professions and age groups.
In the loneliness imposed by the pandemic period and in the clamp of written communication imposed by other social media platforms, the superior aspects of the application are that it provides the two-way audio chat environment that people long for, on many different topics, with many different people, without prior planning, being lively, interactive and open to surprises. With these superiorities, the effect created by Clubhouse raises questions in the minds of researchers such as: “Is this platform and similar ones the future of social media?”, “Is Clubhouse a fashion?”, “What are the attitudes of other social media users towards the platform?”, “For which platforms could it be a threat?”.
In this study which was designed as a preliminary research to answer these questions, it is aimed to reveal how Clubhouse (CH), a new and different social media platform (SMP), is evaluated by Twitter and Sourtimes users, which are powerful and effective SMPs, by conducting sentiment analysis with preliminary data. Thus, it is aimed to reveal the user attitudes towards this SMP using sentiment analysis on big data while Clubhouse is still at the beginning of the ladder. This is one of the leading and largest studies on sentiments towards Clubhouse.
In this chapter, firstly, brief information will be given about the concept of social media, Twitter, and Sourtimes, which are chosen as the research platforms, and Clubhouse, which is determined as the research topic. Then, the concept of sentiment analysis is briefly explained, and the results of the sentiment analysis on all Turkish tweets containing the word “Clubhouse” and all entries in the Clubhouse heading in the Sourtimes between July 1, 2020, and July 1, 2021, were presented and examined.
Key Terms in this Chapter
Hugging Face: Hugging Face is the #1 NLP startup, with over a thousand firms, including Bing, Apple, and Monzo, using the library in their production.
Clubhouse: Clubhouse is a new audio-based social media form with invitation-based participation, where people can chat instantly on any topic.
Sentiment Analysis: Sentiment Analysis is a sub-research area of Natural Language Processing (NLP) studies. With sentiment analysis, also known as Opinion Mining, subjective language elements in oral or written texts about a particular subject can be interpreted.
Unsupervised Learning: Unsupervised learning is the approach used when little or no idea of what the desired output from the data looks like. Various models and structures can be created by clustering the data based on the relationships between the variables.
Social media: Users construct online communities to exchange information, ideas, career, interests, personal messages, and other content via social media (such as websites for social networking and microblogging) (such as videos).
Text Mining: Text mining, which is a special area of data mining, is the application of data mining methods on texts written or shared by people with the help of natural language processing techniques. Many methods have already been developed on the analysis of texts, identifying important terms, summarizing, classifying, or grouping texts.
Supervised Learning: A supervised learning technique uses a known set of input data and known responses to train a model to make credible predictions for new data.
Sourtimes: Sourtimes is a participatory dictionary containing comments from registered authors.