Acomputerprogrammehasbeenusedtopredicthowpeoplearefeelingwhentheytype.
Inastudyparticipantswereaskedtotypeaparticularphraseandtheprogrammethenestimatediftheywerehappy, sadandsoon.
Remarkablyitwascorrect 70 percentofthetime, andthefindingscouldleadtosmarterartificialintelligenceinthefuture.
ThestudypublishedinthejournalBehaviourandInformationTechnologywascarriedoutbyresearchersattheIslamicUniversityofTechnologyinBangladesh, reportsLiveScience.
Intheresearch 25 peoplerangingfrom 15 to 40 yearsoldwereaskedtoretypetwoparagraphsfromAlice'sAdventuresinWonderlandbyLewisCarroll.
Thisallowedtheprogrammeanopportunitytounderstandhowtheiremotionschangedtheirtypingstyle.
Moodstestedwerejoy, fear, anger, sadness, disgust, shameandguilt.
Theresultwasthattheprogrammecouldidentifyaperson'smoodcorrectly 70 percentofthetime.
Itwasmostsuccessfulinidentifyingjoyfulnesswithan 87 percentsuccessrating, whilesadnesswastheleastaccurateat 60 percent.
Theresearchersnotedthattheparticipantswerelesslikelytoinputdatawhentheywereinabadmood, whichmayexplaintheinaccuracyofthelatterresult.
据英国《每日邮报》10月15日报道,一款计算机程序可以预测出人们打字时的情绪。
在一项研究中,参与者被要求输入特定的短语,然后该程序将预测他们是高兴、难过还是其它情绪。