Cornell University researchers have examined more than 500 million Twitter posts from 2.5 million users across 84 countries over two years to find out how people's moods differ through the day.

The study reveals that across continents and cultures, people generally wake up in a good mood but get cranky toward the end of the day, researchers said Thursday.

The study by Cornell graduate student Scott Golder was published in the journal Science Thursday.

Golder, with another graduate student, pulled up information from Twitter accounts from February 2008 through January 2010 with the help of a computer program.

After gathering information from tweets, Golder and his academic advisor, Michael Macy, put the messages into a text analysis program, the Linguistic Inquiry and Word Count, which helped them find out how many times users posted words connected with positive or negative thoughts.

The text analysis program also helped them to find out the time the tweets were posted and the language used.

The researchers classified the words into positive and negative categories.

Words like awesome, certainly, fine, fondly, outgoing, paradise, pleasing, pretty, special, sunny, thanks, fantastic, humor, reward, super, virtue, are positive words which describe good mood, Golder wrote. And words like aching, afraid, cynic, fear, fury, mad, raging, sad, ticked, war, abandon, enemy, panic, remorse, vulnerable, are negative words which describe bad mood.

Macy and Golder also wrote that people tend to follow similar mood trends at specific times of the day, and even over seasons.

They also found that on weekends most people across the world have a good mood until late in the morning, most probably because people were sleeping late.

The study also notes that the mood cycle on weekdays was very similar to that of weekends. which indicates that sleeping patterns and circadian rhythms play a big role.

Researchers also discovered that both positive and negative tweets are common during the evening hours, the most emotionally charged time of the day.

This is a significant finding because one explanation out there for the pattern was just that people hate going to work, Golder told The New York Times. But if that were the case, the pattern should be different on the weekends, and it's not. That suggests that something more fundamental is driving this - that it's due to biological or circadian factors.