Page - 328 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

and thegenderof theauthors in theFacebookcommentsbasedon theirwriting style[8], where amethod is proposed for the identification of the emotions of the texts written andpublishedonFacebook,and thewaypeopleuse language.Theyalsoused themodel to identify the genre of the authors. Another study on emotional analysis was the one executedbyFacebook,manipulating thenewssectionof its socialnetwork[9]. There is still scope for further research into understanding how emotional trajec- tories of different user groups [6] can be affected by location, activity, interaction and other contextual features pertaining to urban areas. These crowdsourced studies can be further supportedby theavaliabilityofseveraldatavisualisationandmapping tools such as openStreetmap [10], an open sourceAPI to generate and share free geographic data usingmapshavebeen reportedon the literature; however little researchhas been found onspatial distributionofemotions. Aswehaveseen in thestateof theart, there is agrowingproblemregardingsuicide in urban areas in Mexico. There is a need determine new ways in which this social problemcanbe identifiedandaffectivelyaddressed throughbetter tools formapping the distributionofwell-beingoverametropolitanarea. Inthispaperwepresentanemotionalreal-timemapoptimizedtosearchforinforma- tionofdifferent typesofusers, indifferent geographical areas,within zonesofdifferent sizes. Figure3(B) showsastratificationofanareaof1,220squarekm.Both thedistances and thenumberofquadrantsof thegridcanbemodified, lookingfor information inspe- cific areas. In addition, our database stores information about registeredusers, allowing us to do, for example, searches of people according to age, gender,workers in thefield ofhealthandeducation. In the experiments presented, data is collected for understanding various activities and emotions affecting university students between 18 and 25 years old. The data is collectedfromstudentsas theyinteractwithapurposebuiltmobileapplicationonadaily basis.Thedata tracks their interactionswithinawiderarea, specifically locationswithin the city of Leon, Guanajuato, Mexico and within a more restricted area, specifically Instituto Tecnolo´gico de Leo´n (ITL), a university located in the city. Using this data emotionalmapsof thecityandcampuscanbecreated identifying themostpredominant emotionsbyzones. This paper is ordered as follows: in the section 2wewill see themost important terms for thiswork. In the section3,weshowanexampleofhowToneAnalyzerworks, howour architectureworks to detect emotional changes and how these changes can be converted tovaluesof energyandpleasantness in theRULERquadrant. In the section4 wedescribe theexperimentperformedfor thedatacollectionandin thesection6wewill see 3 statistical analyzes, two in the city of Leo´nGto,Me´xico and one in a university campus.Finally,wepresentourwork in the future. 2. Background 2.1. SmartCity The term smart cities has emerged during the last decade as a fusion of ideas on how informationandcommunication technologiescould improve the functioningofcitiesby J.J.ManriquezSantosetal. /ACrowdsourcingBasedComputationalMethodology328