How did People Feel about iPhone 12 after the Apple Event? — Sentiment Analysis of Tweets
People are saying that iPhone 12 is disappointing, but is that the majority of them?
What words did people use when they tweet #iphone12? And what feelings did they express?
It’s Apple’s new iPhone release day. While people are eager to find out if the new products are as great as advertised, I’m more curious to find out whether Tweets could reflect the selling performance on the iPhone 12.
You might keep seeing Tweets talking about good/bad things about iPhone 12, but have you ever thought about this — social media could trick you. The contents you see might be what they want you to see after complex algorithms, potentially misleading your feeling about the product.
The good news is, data can help!
After the Apple event, I scrapped 3,200 Tweets with #iphone12 and did sentiment analysis.
Let’s take a look at the data!
First, I used twitteR package in R to scrap 3200 English Tweets with #iphone12.
searchString <- "#iphone12"
numberOfTweets <- 3200
tweets <- searchTwitter(searchString, n = numberOfTweets, lang="en")
tweetsDF <- twListToDF(tweets)And the following text is what it looks like:
Second, I dropped the words that might create a bias to the result:
- Commercial post
- Sweepstakes post
- “iPhones” — it’s the topic, so it can not bring much insight.
drop_keyword <- c("sweepstakes","looking","giveaway","save12hkyouths","chance","win","attexplore","iblason","@TMobile","@Sprint","spectrum",'apple','iphone12','rt','iphone','ar72014','iphone12pro','iphone12promax','iphone12mini','iphone11','iphone11pro','iphone11promax','iphonexs','iphonexsmax')Exploratory Data analysis
The data is cleaner now. Let’s take a look at the word count:
We can see that the most frequent word is “new.” If this is the idea Apple wants to deliver during the event, then it’s pretty successful. On the other hand, if the values that Apple likes the audiences to receive, such as “camera” and “environmental,” are not among the list above, they could consider adjusting some part next time.
Sentiment Analysis
Now it’s time to see how people “feel” about #iphone12!
From the graph above, we can see people are expressing various feelings about the iPhone 12.
In general, positive Tweets are more than negative.
Does this indicate a good selling performance? We can only know when Apple releases the reports.
From this graph above, we can have a better understanding of the reasons behind certain feelings. Did you find anything not what you expected?😁
Finally, I chose three words and find a correlation of words being used in the same tweet.
- “Charger”
- “Magsafe”
- “Verizon” — it seems like most tweets are talking about the upcoming League of Legends mobile game.
Potential bias of this analysis
⚠️To be clear, this whole sentiment analysis is not perfect mainly because:
- It can’t understand sarcastic posts.
- The sample is not fully random. It’s only select the posts on Twitter, and English only. Also, it’s only about 2000 Tweets after data cleaning. This sample can not represent all the opinions online.
If you have any ideas or feedback you would like to share with me, feel free to send a message to me on Linkedin!
