US researchers have created an automated text analysis system to identify features which make a tweet more likely to get attention.
Using a computer algorithm, Cornell University has launched a website to predict which version of a tweet will be retweeted more.
Try it here.
The research might be applied to longer written forms such as essays or how to get an idea accepted in a meeting.
“We’re looking at persuasion everywhere,” says Lillian Lee, professor of computer science and information science.
The researchers collected and compared thousands of tweets from people who tweeted on the same topic more than once but with different wording.
Lee, graduate student Chenhao Tan and Google researcher Bo Pang, reported their results in the journal Proceedings of the Association for Computational Linguistics.
The computer looks for the occurrence of certain keywords and compares “bigrams” or combinations of two words.
Just asking people to share gets results. Words such as “please,” “pls,” “plz” and “retweet” are common in successful messages.
As a general rule, the researchers recommend imitating the style of newspaper headlines.
- Use words which appear often in other retweeted messages.
- Express positive or negative sentiment.
- Refer to other people, not just yourself. Use third person pronouns.
- Use generalisations. Statements which can be applied to a variety of situations are the most often repeated.
- Make it easy to read.
“We would love to capture amusingness or cleverness, but we haven’t found a way to do that yet,” Lee says.
NOW WATCH: Briefing videos
Business Insider Emails & Alerts
Site highlights each day to your inbox.