There are more than 48 million tweets in the dataset, which language is in the ten most using languages. About 1,114,859 tweets mentioned drinks that are about 0.232 per cent of all the tweets in specific languages.
One of the most significant findings is that the number of tweets about alcohol is 0, which means Twitter excludes any tweets related to the alcoholic beverage. In this way, Twitter protects teenagers from drinking alcohol and prevents alcoholic beverage companies from getting valuable data about selling alcohol.
From the bar chart above, we can get some conclusions. Among the tweets about drinking, coffee is the most frequently mentioned drink type, occupying about 40 per cent. The second type is the healthy drink, about two-thirds of the number of coffee. The number of tweets about tea is 211,224, which is about 57,000 langer than the soft drink. Surprisingly, tea is more popular than soft drinks on Twitter.
For the soft drink, people who speak Spanish talked more about it than other languages. On the contrary, Russian least like to drink soft drink. We suggest that if the beverage company wants to advertise their soft drink on Twitter, they should more focus on users who speak Spanish, Portuguese, Turkish, and Italian. Besides, advertising soft drink in Russian on Twitter is not recommended, based on data.
On Twitter, Indonesian users talked more about healthy drink than other language users. Therefore, the beverage company should spend more money on advertising healthy drink in Indonesian, such as milk, juice and chocolate. User in Russian, German and Dutch are less interested in the healthy drink. Hence, the beverage company should reduce advertising expenditure in these fields.
Coffee and tea are two types of drinking which help people focus and stay alert. Hence, we compare coffee and tea together in different language users. From the bar graph below, we can illustrate that, except for the Turkish users, other language users like coffee more than tea. And Italian users like coffee most, almost two times the percentage of other language users. From this graph, Turkish users are the most potential customers for tea, while Italian users are the most worthy customers for coffee.
From that line chart below, we can find that the percentage of people talked about healthy drink, soft drink, and tea maintain stable. Although tea also has the effect of keeping energy and productivity like coffee, tea drinkers do not drink tea in the morning for its effect. Unlike tea, the percentage of people who mentioned coffee peaks at eight o'clock, which means people drink coffee mainly to focus on their study or work. As a result, the coffee advertisements will be more effective in the morning like the morning news and papers.
To summarize, we dig out the value of the Twitter dataset about the beverage using Big Data. Apart from the alcohol that Twitter filters, we get the conclusions about coffee, tea, healthy drink, and soft drink and give some suggestions to the beverage company based on this dataset.