University researchers have successfully analyzed 361 billion words.
A study by a team of UVM mathematicians found that positive words far outnumber negative words in the English language, according to University Communications.
“It is exciting that data from online, public communications offers this grand opportunity to observe and explore the nature of humans and their behaviors,” said Isabel Klouman, a UVM alumni and team member. The results of their study have many anthropological and sociological implications, she said.
“This result has implications for behavioral economics, sociology, and the study of language evolution in cultures,” Kloumann said. “We are making the data publicly available, and it will likely be used by linguists, computer scientists, economists, and business people.” The team began working on the study in the fall of 2011 as part of UVM’s Complex Systems Center, and they used the supercomputer at the Vermont Advanced Computing Center, she said.
They covered millions of pages of material spanning from books to tweets and newspapers to song lyrics. They took the top 5,000 most commonly used words from each of those four areas, then combined those lists to form a list of 10,222 words.
“Observing people’s online communication gives you insight into their daily behaviors and mood,” Kloumann said. “Sure the metric isn’t flawless, but it’s a very good start and we’ll continue to refine it,”
First-year Emily Sleeper said that she was intrigued by the findings of the study.
“I think it’s surprising [that positive language outnumbers the negative] because the English language is so full of slang with inappropriate and negative connotations,” Sleeper said. Kloumann said her team wasn’t necessarily expecting the results of the study.
“We think the result is very interesting, and that if this positivity bias is universal across cultures and languages, it might be a reflection of humans’ pro-social natures and the corresponding influence on the evolution of language and communication, ” she said.
Those interested in learning more about this research can check out the applied math group’s blog at: onehappybird.com. Isabel Kloumann will be speaking about this work at the UVM TEDx conference “Big Data, Big Ideas” on October 28th.
