El discurso de odio en Twitter durante la crisis migratoria de Ceuta en mayo de 2021

ICONO 14, Re is a de comunicación y ecnologías eme gen es, 20(2), 2022. ISSN: 1697-8293 h ps://doi.o g/10.7195/ i14. 20i2.1858 ICONO 14 es edi ada po : F ancisco Ga cía (Uni e sidad Complu ense de Mad id, España) y Manuel Ge udix (Uni e sidad Rey Juan Ca los (España) bajo una licencia de licencia de C ea i e Commons Reconocimien o 4.0 In e nacional. Monog áficos Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 El discu so de odio en Twi e du an e la c isis mig a o ia de Ceu a en mayo de 2021 Discu so de ódio no Twi e du an e a c ise mig a ó ia de Ceu a, em Maio de 2021 A ánzazu Román-San-Miguel1* F ancisco J. Oli a es-Ga cía1** Salud Ma ía Jiménez-Za a2*** 1 Depa men o Jou nalism II, Facul y o Communica ion, Uni e sidad de Se illa, Spain 2 Depa men o Compu e Science, Uni e sidad de Jaén, Spain * Associa e P o esso (Depa men o Jou nalism II). Facul y o Communica ion, Uni e sidad de Se illa, Spain ** Associa e P o esso (Depa men o Jou nalism II). Facul y o Communica ion, Uni e sidad de Se illa, Spain *** P o esso (Depa men o Compu e Science). Uni e sidad de Jaén, Spain Recei ed: 23/02/2022; Re ised: 19/03/2022; Accep ed: 29/06/2022; Published: 28/07/2022 T ansla ion o English: ALTALINGUA, S.L. To ci e his a icle: Román-San-Miguel, A ánzazu; Oli a es-Ga cía, F ancisco J.; & Jiménez-Za a, Salud Ma ía. (2022). Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 ICONO 14. Scien ific Jou nal o Communica ion and Eme ging Technologies, 20(2). h ps://doi.o g/10.7195/ i14. 20i2.1858 Abs ac 2Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 This pape analyses ha e speech wi hin messages pos ed on Twi e om 17-25 May 2021 du ing he c isis caused by housands o immig an s en e ing on he Ta ajal bo de in Ceu a. The aim o he esea ch was o g oup messages ha included ha e speech in o hemes. To do his, a mixed me hodology was used, and esul ed in six hemes being iden ified. O hese, ou ocused on poli ical issues, accoun ing o 80% o he messages, and only 20% e e ed o acism and immig a ion. Up o fi e campaigns o unknown o igin ha e also been iden ified. This pape concludes ha ha e speech ocuses mo e on poli ical issues han on he p oblem o immig a ion i sel , i s causes, consequences and possible solu ions. Keywo ds: Ha e speech; Twi e ; Ceu a; social media; online campaigns; immig a ion Resumen Es e abajo analiza el discu so de odio en los mensajes publicados en Twi e desde el 17 al 25 de mayo de 2021 du an e la c isis p oducida po la en ada de miles de inmig an es en la on e a del Ta ajal en Ceu a. El obje i o de la in es igación es ealiza una clasificación emá ica de los mensajes que incluyen discu so de odio. Pa a ello, se ha empleado una me odología mix a y como esul ado se han podido di e encia seis emas, de los cuales cua o se cen an en emas polí icos, suponiendo el 80% de los mensajes, y solo el 20% de ellos se efie en a acismo e inmig ación. Además, se han de ec ado has a cinco campañas de o igen desconocido. Es e abajo concluye que los discu sos de odio se cen an más en emas polí icos que en la p opia p oblemá ica de la inmig ación, sus causas, sus consecuencias y las posibles soluciones. Palab as cla e: Discu so de odio; Twi e ; Ceu a; edes sociales; campañas online; inmig ación Resumo Es e documen o analisa o discu so do ódio nas mensagens publicadas no Twi e de 17 a 25 de Maio de 2021 du an e a c ise causada pela en ada de milha es de imig an es na on ei a de Ta ajal em Ceu a. O objec i o da in es igação é le a a cabo uma classificação emá ica das mensagens que incluem o discu so do ódio. Pa a es e fim, oi u ilizada uma me odologia mis a, e como esul ado, o am iden ificados seis emas, qua o dos quais se concen am em ques ões polí icas, ep esen ando 80% das mensagens, e apenas 20% das quais se e e em ao acismo e à imig ação. Além disso, o am de ec adas a é cinco campanhas de o igem desconhecida. Es e es udo conclui que o discu so do ódio se concen a mais em ques ões polí icas do que no p oblema da imig ação em si, nas suas causas, consequências e possí eis soluções. Pala as-cha e: discu so do ódio; Twi e ; Ceu a; edes sociais; campanhas em linha; imig ação 1. In oduc ion Social media ha e allowed he use o eedom o exp ession o degene a e in o abuse, c ea ing he need o limi s. So much so, ha in ecen yea s he media deba e has ocused on whe he eedom o exp ession is a isk when ce ain speech - especially ha e speech - is being ough in he cou s. As Bu gos-Ga cía (2019) a gues, "I people beha e anonymously Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 3 and ail o adhe e o c i ical, cons uc i e hinking, he limi s on eedom o exp ession a e exceeded o he de imen o he igh s o o he s" (p. 138). Howe e , a dis inc ion should be made be ween diffe en o ms o exp ession ha may be ude, and ha e c imes hemsel es, which lead he legal sys em o place limi s on eedom o exp ession (p. 149). The bi h o social ne wo ks can be aced back o he beginning o he 21s cen u y, al hough hey did no become popula un il yea s la e , when he numbe o use s and he unc ionali ies o hese pla o ms inc eased. Only a decade la e , esea che s such as Falxa (2014) began o s udy ha e speech on social media, claiming ha " he figh agains ha e speech on social media comes up agains nume ous obs acles" (p. 104). Al hough s eps in his di ec ion we e al eady being aken in Eu ope, such as he decision aken on 24 Janua y 2013 by he Judge o Fi s Ins ance o Pa is's T ibunal de G ande Ins ance, equi ing US- based Twi e " o se up an ale mechanism allowing F ench use s o epo any abuse o ha e messages o possible emo al, and o p o ide he F ench au ho i ies wi h de ails o any use s who ha e pos ed illegal wee s" (p. 104). On 31 May 2016, he Eu opean Union equi ed Facebook, Google, Mic oso and Twi e o ag ee o a code o conduc ha would equi e hem o e iew and emo e wi hin 24 hou s any illegal ha e speech pos ed on hei se ices (Bish e al., 2020, p. 244). The scien ific li e a u e app oaches ha e speech om diffe en a eas o s udy - especially law (Howa d, 2019; Rollne -Lie n, 2019; Bau is a-O uño, 2017; Falxa, 2014). Howe e , i is significan ha s udies a e ca ied ou as pa o communica ion sciences - especially when he e a e hose who conside social ne wo ks o be media (Hü -He e a, 2012; Pan oja- Cha es, 2011). This is a poin o iew ha can be dispu ed. Howe e , wha is no dispu ed is ha messages pos ed on social media each millions o people e e y day and i is impo an o know whe he , as channels o communica ion, hey sp ead ha e c imes. On his, a ecen s udy by A cila-Calde ón e al. (2022, p. 32) ound ha , "in Eu opean egions whe e he e is a highe p opo ion o immig an s, hey ecei e g ea e public suppo " and, he e o e, " egions in which suppo was g ea e saw a lowe le el o ha e speech on Twi e ". Howe e , as his pape shows, he e is s ill ha e c ime on social ne wo ks such as Twi e which we should wo k o e adica e. Some s udies abou social media ha ha e been published ha e ocused on ha e c imes in ega d o gende o sexual o ien a ion - such as A cila-Calde ón e al. (2021a). The au ho s also p esen " he fi s p o o ype o au oma ically de ec ing ha e speech on Twi e in Spanish, specifically we e i is mo i a ed by gende and sexual o ien a ion". This esea ch is based on e en s in Spain on 17-18 May 2021, when housands o people en e ed he coun y h ough he bo de wi h Mo occo a El Ta ajal (Ceu a). This p omp ed Spain’s P ime Minis e , Ped o Sánchez, o isi he a ea on 18 May, and on he same day Reu e s p o ided he media wi h an image o a Spanish Red C oss olun ee com o ing a sub-Saha an immig an , which gene a ed nume ous nega i e wee s. 4Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 The igge o Mo occo allowing housands o people o en e Spain was he Polisa io F on leade , B ahim Ghali, being admi ed o a Spanish hospi al on an Alge ian passpo , wi hou p io no ice being gi en o he Mo occan au ho i ies, a a ime o g ea ension, as in No embe 2020 hea y clashes had e up ed be ween Mo occo and he Polisa io F on o e con ol o Wes e n Saha a. All o his p o oked majo eac ion on social ne wo ks, bo h in Spain and Mo occo. Some use s ook ad an age o he occasion o pos xenophobic, iolen , acis o ha e speech messages. Immig a ion has always been a ocus o omen ing speech a ound ejec ion and ha ed. Acco dingly, we can ci e s udies on he ha ed ex olled by na ionalism in he Uni ed S a es, wi h mig an s being posi ioned as a h ea (Mo gen eld, 2016); and Ge many, whe e he Al e na i e o Ge many (A D) pa y - which calls i sel an i- e ugee and an i-immig a ion - became he hi d s onges pa y in he Ge man pa liamen in 2017 (Valdez-Apolo e al., 2019, p. 365). The la e s udy concludes ha he p edominan one on Twi e is nega i e, which he au ho s a ibu e o he ac ha immig a ion is equen ly ela ed o "hos ile e en s o ci cums ances o a bellicose, economic, poli ical, demog aphic and/o clima ic na u e" (p. 379). Howe e , i has he limi a ion ha " he s udy was done by manually analysing con en " (p. 385). As we will see below, his a icle has used a mixed me hod o he analysis. Ano he ocus o ha e ul and iolen publica ions is Islamophobia (Awan, 2014, Zamo a e al., 2021). 1.1. F eedom o exp ession and social media The ac ha eedom o exp ession is a igh does no mean ha i is wi hou a numbe o limi a ions ha ha e been legally pu in place o e he yea s. These limi a ions a e un ela ed o he exp ession o ideas ha a e pe cei ed as nega i e - as, i his we e he case, plu alism, ole ance and b oadmindedness in a democ a ic socie y would disappea (González-He e a, 2018, p. 3). On his, he media a e pa icula ly p o ec ed as in o me s and opinion o me s, e en whe e his igh has no always been exe cised in an e hical manne . By way o example, González-He e a ci es he alse in o ma ion published in he UK du ing he campaign leading up o he e e endum o lea e he Eu opean Union and Donald T ump's 2016 p esiden ial elec ion. Acco ding o some s udies, he US p esiden had shown himsel o be 'an in e e a e lia , immune and defian o ac - checking' (Eche a ía, 2016, p. 10). I was no he media, bu social ne wo ks ha sp ead he mos ake news du ing he US Republican candida e's campaign, which was accen ua ed du ing he Co id-19 pandemic - ake news ha was hen passed on o he media ha used social ne wo ks as sou ces o in o ma ion (Fe nández-Muñoz e al., 2022; Gu ié ez-Vid io, 2020; Pé ez-Cu iel and Limón- Naha o, 2019). Indeed, social media and especially Twi e had o in e ene o cu b Donald Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 5 T ump's misuse o he pla o m, especially a e he assaul on he US Capi ol on 6 Janua y 2021 and his alse messages, i no inci ing ha ed and iolence. On 8 Janua y, Twi e shu down Donald T ump's accoun and la e Reddi shu down he accoun o T ump suppo e s o p omo ing ha e ul beha iou , ollowed by Twi ch shu ing down T ump's accoun . On Twi e 's help websi e he e is an en i e sec ion dedica ed o he company's policies whe e ha e speech is de ec ed in pos ed con en (Twi e , 2021). Du ing he cou se o his in es iga ion, se e al accoun s we e ound o ha e been dele ed o pos ing ha e speech con en on a numbe o occasions. 1.2. Ha e speech on Twi e Ha e is deemed o be ep ehensible conduc in he mo al, eligious, social and legal sphe es, as i can ha m igh s such as a pe son o g oup's honou o digni y (Mo e ón- Toque o, 2012, p. 4). As ea ly as 2012, an a emp was made o gi e a name o ha e- mo i a ed beha iou on he in e ne , which Mo e ón-Toque o (2012) called "cybe ha e" - e en hough, in his s udy, he in e ne is e e ed o in a e y gene ic way, wi hou s udying he impac o social media. Wha i highligh ed was he difficul y o legisla ing in such an open and in e na ional domain like he in e ne , bu also he ea ha i ialising his issue in a medium accessed by young people as could be a oused by a mo e influen ial public (p. 15). To define which ma e ial is acis and xenophobic, we e e ed o he Addi ional P o ocol o he Con en ion on Cybe c ime, conce ning he c iminalisa ion o acis and xenophobic ac s commi ed h ough compu e sys ems, which was ag eed in S asbou g on 28 Janua y 2003 and published in Spain's Official Go e nmen Gaze e Numbe 26, Sec ion 1 on 30 Janua y 2015: "Racis and xenophobic ma e ial" means any w i en ma e ial, images o any o he ep esen a ion o ideas o heo ies ha ad oca es, p omo es o inci es ha ed, disc imina ion o iolence, agains any pe son o g oup o pe sons, on he g ounds o ace, colou , descen o na ional o e hnic o igin, as well as eligion inso a as eligion is used as a p e ex o any o hese ac o s. (p. 7.216) The Ha e C ime Su ey Repo o he pe iod Decembe 2020 o Ma ch 2021, by he Sec e a y o S a e o Secu i y o Spain's Home Office (López-Gu ié ez e al., 2021), classifies ha e c ime as: acism/xenophobia, sexual o ien a ion/ LGBTphobia, eligion, ideology and disabili y/illness. All hese mani es a ions o ha e could be ound among he messages downloaded om Twi e o his p ojec . Acco ding o a epo published by Ba celona Ci y Council, as ea ly as 2014 he e we e a ound 10,000 wee s a day ha included acis insul s in English - equi alen o one in e e y 6Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 15,000 wee s (Cabo-Isasi and Ga cía-Juana ey 2017, p. 5). Yea s la e , he media would echo Twi e 's decision o dele e mo e han a housand wee s in which acis abuse was de ec ed and i was s a ed ha he B i ish go e nmen was wo king on laws in his a ea (Cano, 2021). We ound some ecen s udies abou ha e speech on Twi e , such as ha o Valdez- Apolo e al. (2019), who claim ha he p e ailing sen imen in messages abou immig an s is nega i e, and mo e so han abou e ugees. Those au ho s conside ha his "nega i e one" may ha e i s o igin in he ac ha " he mig a ion phenomenon" in he media is "nega i e in na u e", as i is equen ly " ela ed o hos ile e en s o ci cums ances o a bellicose, financial, poli ical, demog aphic and/o clima ic na u e" (p. 379). They also add ess he s igma isa ion o e ugees h ough ha e messages on social media F ías and Seoane (2019), and A cila-Calde ón e al. (2021b). This is some hing ha can be seen in a la en o m in he s udy a hand, as he massi e en y o immig an s in o a coun y - and encou aged by a neighbou ing coun y's au ho i ies' laxi y - could be seen by he ecei ing popula ion as a h ea o i s in eg i y as a coun y. Fo , al hough online ha e speech migh ini ially be hough o cause no ha m due o i s i ual na u e, s udies by Williams e al. (2020) ha e shown a link be ween ha e speech on social media and ac ual c ime - especially agains black and Muslim communi ies. O he s udies ha e app oached he e ugee issue by disagg ega ing he s udy by gende , such as Amo es e al. (2020), who conclude ha he image o e ugee women is as passi e agen s in he images published by he media. Poli ical discou se has also been ela ed in diffe en s udies o ha e messages ia Twi e (A cila-Calde ón e al., 2021c; A cila-Calde ón e al., 2020; Case o-Ripollés e al., 2020; Paz, 2021). 2. Ma e ial and me hods A mixed me hodology was used o his p ojec - e y simila o ha seen in o he p ojec s ela ed o ha e speech and Twi e . Fo example, in Mi ó-Lina es (2016), in which bo h quan i a i e analysis wi h da a mining echniques and quali a i e analysis o con en a e used. Acco dingly, echniques ha au oma ically fil e Twi e con en based on he use o specific lexicons ha e also been used - as can be seen in Jiménez-Za a e al. (2021). Fi s , all messages published in Spanish be ween 17-25 May 2021 con aining he wo ds Ceu a and/o Mo occo we e downloaded using Twi e 's API V2. An addi ional download was also done based on a se ies o hash ags, as shown in Table 1. The lis o ags was ob ained using he T endinalia, Accoun Analysis and Google T ends websi es, based on Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 7 he opics ha mos in e es ed use s and he ags used in he messages. This was hen all con e ed in o a sp eadshee . A o al o 310,907 wee s we e downloaded. Many o hem we e e wee ed pos s - ha is, o iginal pos s ha o he people simply copied and pas ed as hey we e ia hei own accoun s, wi hou con ibu ing any con en . Thus, hese messages needed o be elimina ed, lea ing 38,901 pos s a he end o his p ocess. Table 1 shows he numbe o messages e ie ed acco ding o keywo d and ag classifica ion: Table I. Comple e lis o downloaded messages o ganised by keywo d and ag Sou ce: Au ho 's own A second fil e ing was hen pe o med o emo e duplica e en ies, as a wee can include se e al keywo ds and ags a he same ime. A e he new fil e ing p ocess, he numbe o publica ions was educed o 36,887. In o al, 18,441 o iginal messages we e iden ified, he es being hose ha we e quo ed (2,932) o eplied o (15,514). Fo he quan i a i e analysis, ou specific lexicons we e used ha can be help ul in de ec ing ha e speech. These lexicons ha e been sa is ac o ily e alua ed in he s udy by Plaza-del-A co e al. (2020). The numbe o wo ds con ained in each lexicon along wi h a couple o significan examples a e shown below: 8Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 Xenophobia: 44 wo ds (e.g. Moo /Muslim), nigge ) Immig a ion: 250 wo ds (e.g. depo him, e ugee) Misogyny: 183 wo ds (e.g. bi ch, slu ) Insul s: 279 (e.g. bas a d, aggo ) To acili a e he analysis, all in o ma ion was o ganised in a sp eadshee consis ing o 18 fields. The fi s en fields we e di ec ly ela ed o he in o ma ion ex ac ed ia he Twi e API; he o he eigh we e c ea ed based on he con en p o ided by he ou lexicons. Each o hese lexicons gene a ed wo fields in he sp eadshee : one indica ing he p esence (o no ) o wo ds included wi hin hem and he o he con aining he e ms ha we e ound: Id: Unique iden ifie o he wee . Tex : Message w i en in he wee . Topic: Field au oma ically gene a ed based on use o he lexicon fil e , which classifies messages acco ding o whe he hey deal wi h immig a ion, xenophobia, insul s o misogyny. The esul o his classifica ion can be seen in Table 2. Type: O iginal, eply o quo e, depending on whe he he message is an o iginal pos o whe he i is a commen on o quo ed om ano he . Sensi i e con en : Field p o ided au oma ically by Twi e , showing ei he ue o alse. Re wee coun : Numbe o imes an o iginal pos has been e wee ed. Commen coun : To al numbe o commen s ecei ed o each wee . Quo ed coun : Numbe o imes a use has quo ed a message. Likes coun : To al numbe o use s who ha e liked a message. These ou fields ha e been used o assess he popula i y o wee s. Sou ce: The o igin o he publica ion, depending on whe he i was wee ed om a mobile o desk op applica ion. To al numbe o wo ds on immig a ion: Numbe o wo ds in he wee ha a e included in he immig a ion lexicon. Numbe o wo ds on immig a ion: Numbe o wo ds om he immig a ion lexicon ha a e used in he wee . To al numbe o insul s: Numbe o lexicon insul s used in he wee . Insul s: Wo ds om he lexicon ela ed o insul s and used in he wee . To al numbe o misogynis ic wo ds: Numbe o wo ds om he misogyny lexicon ha a e used in he wee . Misogynis ic wo ds: Wo ds om he misogyny lexicon ha a e used in he wee . To al numbe o xenophobic wo ds: Numbe o xenophobic wo ds used in he wee . Xenophobic wo ds: Wo ds om he xenophobia lexicon ha a e used in he wee . Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 9 Table II. Themed classifica ion based on he lexicon fil e s Sou ce: Au ho 's own Following his au oma ed o ganisa ion, he con en was analysed manually by eading he messages ma ked by he lexicon heme fil e , along wi h all he messages wi h hash ags. Thus, o he 36,887 wee s downloaded wi hou duplica es, a o al o 7000 wee s suspec ed o con aining ha e speech we e manually e iewed. O hese, 4964 include one o he wo ds in he lexicons, while 1817 ela e o he sum o all messages ha included one o he hash ags. The ac ha a pos has a wo d om a lexicon does no imply ha he wee is iolen , jus as he e a e messages ha a e no ma ked by lexicons ha include ha e ul con en . To find he la ges numbe o hese messages, an addi ional sea ch o he wo ds wa , in asion, bo de , a my, humilia ion, dea h and expulsion was done. This added 219 messages o hose analysed, b inging he o al o 7000. These messages we e manually e iewed and 1000 wee s we e ound ha ac ually included ha e speech. Among he es , al hough wo ds we e used ha could indica e ha ed, hey we e no in he con ex used. The exac numbe was 1005, bu was educed o 1000 o simpli y he calcula ions. 3. Findings and discussion Based on he analysis o he con en o he messages, hey we e classified in o six hemed a eas. This classifica ion does no ake accoun o messages ha o m pa o o ganised campaigns, which a e discussed in sec ion 3.2. 16 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 Re e ences Amo es, Ja ie J., A cila-Calde ón, Ca los, & González-de-Ga ay, Bea iz (2020). The gende ed ep esen a ion o e ugees using isual ames in he main wes e n eu opean media. Gende Issues, 37(4), 291-314. h ps://doi.o g/10.1007/s12147-020-09248-1 A cila-Calde ón, Ca los, Amo es, Ja ie J., Sánchez-Holgado, Pa icia, & Blanco-He e o, Da id (2021a). Using shallow and deep lea ning o au oma ically de ec ha e mo i a ed by gende and sexual o ien a ion on wi e in spanish. Mul imodal Technologies and In e ac ion, 5(10) h ps://doi.o g/10.3390/m i5100063 A cila-Calde ón, Ca los, Blanco-He e o, Da id, F ías-Vázquez, Maximiliano, & Seoane-Pé ez, F ancisco (2021b). Re ugees welcome? online ha e speech and sen imen s in wi e in spain du ing he ecep ion o he boa aqua ius. Sus ainabili y (Swi ze land), 13(5), 1-17. h ps://doi. o g/10.3390/su13052728 A cila-Calde ón, Ca los, Blanco-He e o, Da id, Ma siola, Ma ía, Olle -Alonso, Ma ín, Sa idou, Theodo a, Splendo e, Se gio, & Veglis, And eas (2021c). F aming mig a ion in sou he n eu opean media: Pe cep ions o spanish, i alian, and g eek specialized jou nalis s. Jou nalism P ac ice, h ps://doi.o g/10.1080/17512786.2021.2014347 A cila-Calde ón, Ca los, de la Vega, Gonzalo, & Blanco He e o, Da id (2020). Topic modeling and cha ac e iza ion o ha e speech agains immig an s on wi e a ound he eme gence o a a - igh pa y in Spain. Social Sciences, 9(11), 1-19. h ps://doi.o g/10.3390/socsci9110188 A cila-Calde ón, Ca los, Sánchez-Holgado, Pa icia, Quin ana-Mo eno, C is ina, Amo es, Ja ie J., & Blanco-He e o, Da id (2022). Ha e speech and social accep ance o mig an s in eu ope: Analysis o wee s wi h geoloca ion. [Discu so de odio y acep ación social hacia mig an es en Eu opa: Análisis de ui s con geolocalización] Comunica , 30(71), 1-13. h ps://doi.o g/10.3916 /C71-2022-02 Awan, Im an (2014). Islamophobia and Twi e : A ypology o online ha e agains Muslims on social media. Policy & In e ne , 6(2), 133-150. h ps://doi.o g/ggb54 Bau is a-O uño, Rebeca (2017). ¿E es un cibe ha e ? P edic o es de la comunicación iolen a y el discu so del odio en In e ne . In e na ional e-Jou nal o C iminal Science, 11, 1-28. h ps://bi .l y/3kiYL L Bish , Akanksha, Singh, Annapu na, Bhadau ia, H. S., Vi mani, Ji end a, & K i i (2020). De ec ion o Ha e Speech and Offensi e Language in Twi e Da a Using LSTM Model en Jain, Sh u i, & Paul, Sudip (coo d.), Recen T ends in Image and Signal P ocessing in Compu e Vision (pp. 243-264). Sp inge . h ps://doi.o g/h 4 Bu gos-Ga cía, Olga (2019). Los lími es a la libe ad de exp esión: el discu so del odio en In e ne en Ma ín-Conejo, Se gio (ed.). El mundo a a és de las palab as. Lenguaje, géne o y comunicación (pp. 138-150). Dykinson, S.L. h ps://doi.o g/gwbs Cabo-Isasi, Alex, & Ga cía-Juana ey, Ana (2017). El discu so de odio en las edes sociales: un es ado de la cues ión. Ayun amien o de Ba celona. h ps://bi .ly/3B 92 b Cano, Manuela (12 de julio de 2021). In es igan a aques acis as a jugado es ingleses as la final de la Eu ocopa. F ance 24. h ps://bi .ly/2W V3Hq Case o-Ripollés, And eu, Micó-Sanz, Joseph-Lluís., & Díez-Bosch, Mí iam (2020). Digi al public sphe e and geog aphy: The influence o physical loca ion on wi e ’s poli ical con e sa ion. Media and Communica ion, 8(4), 96-106. h ps://doi.o g/10.17645/mac. 8i4.3145 Eche a ía, Bo ja (2016). Más ‘ ac -checking’ con a la pos e dad. Cuade nos de pe iodis as: Re is a de la Asociación de la P ensa de Mad id, 33, 9-16. h ps://bi .ly/3gyBgwL Ha e speech on Twi e du ing he Ceu a mig a ion c isis in May 2021 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 17 BOE (30 de ene o de 2015). Ins umen o de Ra ificación del P o ocolo adicional al Con enio sob e la Cibe delincuencia ela i o a la penalización de ac os de índole acis a y xenó oba come idos po medio de sis emas in o má icos, hecho en Es asbu go el 28 de ene o de 2003. BOE, n. 26, sec. I, 7214-7224. h ps://bi .ly/3mByWsQ Falxa, Joana (2014). Redes sociales y discu sos de odio: un en oque eu opeo en Pé ez Ál a ez, Fe nando (coo d.). Mode no discu so penal y nue as ecnologías (pp. 89-106). Ediciones Uni e sidad de Salamanca. Fe nández-Muñoz, C is óbal, Rubio-Mo aga, Ángel Luis, & Ál a ez-Ri as, Da id (2022). The Mul iplie Effec on he Dissemina ion o False Speeches on Social Ne wo ks: Expe imen du ing he Silly Season in Spain. In: Lahby, Mohamed, Pa han, Al-Sakib Khan, Maleh, Yassine, & Ya ooz, Wael Mohamed Shahe . (eds) Comba ing Fake News wi h Compu a ional In elligence Techniques. S udies in Compu a ional In elligence, ol 1001. Sp inge , Cham. h ps://doi.o g/10.1007/978 -3-030-90087-8_12 F ías Vázquez, Maximiliano, & Seoane Pé ez, F ancisco (2019). Ha e Speech in Spain Agains Aqua ius Re ugees 2018 in Twi e . Pape p esen ed a he ACM In e na ional Con e ence P oceeding Se ies, 906-910. h ps://doi.o g/10.1145/3362789.3362849 González-He e a, Daniel (2018). Libe ad de exp esión y discu so de odio en Eu opa: p o egiendo a las mino ías en iempos de pos e dad en Rod íguez-Ga cía, Nicolás, Ca izo-González-Cas ell, Adán, & Le u ia-In an e, F ancisco J. (coo d.) Jus icia Penal Pública y Medios de Comunicación (pp. 549-573). Ti an lo Blanch. Gu ié ez-Vid io, Sil ia (2020). El discu so polí ico en la e a digi al. Donald T ump y su uso de Twi e . Es udios del discu so, 6(1), 56-81. h ps://bi .ly/3lByCIm Howa d, Jeff ey W. (2019). F ee Speech and Ha e Speech. Annual Re iew o Poli ical Science, 22, 93-109. h ps://doi.o g/gh 2 Hü -He e a, Ha old (2012). Las edes sociales: Una nue a he amien a de di usión. Reflexiones, 91(2), 121-128. h ps://bi .ly/3EyaONC López-Gu ié ez, Ja ie , Fe nández-Villazala, Tomás, Máñez-Co inas, Ca los, San-Abela do-An a, Ma ía Yami , Gómez-Es eban, Jesús, Sánchez-Jiménez, F ancisco, He e a-Sánchez, Da id, Ma ínez-Mo eno, F ancisco, Rubio-Ga cía, Ma cos, Gil-Pé ez, Vic o ia, San iago-O ozco, Ana Ma ía, & Gómez-Ma ín, Miguel Ángel (2021). In o me de la encues a sob e deli os de odio. Sec e a ía Gene al Técnica del Minis e io del In e io , NIPO: 126 21 071 6. h ps://bi .ly/3nM bMyS Jiménez-Za a, Salud Ma ía, Sáez-Cas illo, An onio José, Conde-Sánchez, An onio, & Ma ín-Valdi ia, Ma ía Te esa (2021). How do sen imen s affec i ali y on Twi e ? Royal Socie y Open Science, 8(4), 1-11. h ps://doi.o g/gwd Maqueda, Ad ián (19 de mayo de 2021). Ni es en Tu quía ni es an igua: la imagen del gua dia ci il en Ceu a esca ando a un bebé es eal. New al. h ps://bi .ly/2XoGqoG Mi ó-Lina es, Fe nando (2016). Taxonomía de la comunicación iolen a y el discu so del odio en In e ne . Re is a d’In e ne , D e i Polí ica, 22, 93-118. h ps://doi.o g/g4b6 Mo e ón-Toque o, Ma ía A ánzazu (2012). El cibe odio, la nue a ca a del mensaje de odio: en e la cibe c iminalidad y la libe ad de exp esión. Re is a ju ídica de Cas illa y León, 27, 1-18. h ps://bi .ly/3DwlLQb Mo gen eld, Leand o (2016). Es ados Unidos: T ump y la eacción xenó oba con a la inmig ación hispana. Conflic o Social, 9(16), 15-33. h ps://bi .ly/3pXix3B Pan oja-Cha es, Aan onio (2011). Los nue os medios de comunicación social: las edes sociales. Tejuelo, 12, 218-226. h ps://bi .ly/2ZAa6QD 18 Icono14 |Julio-Diciemb e 2022 Volumen 20 N° 2 | ISSN: 1697-8293 | DOI: h ps://doi.o g/10.7195/ i14. 20i2.1858 Paz, Ma ía An onia, Mayagoi ia-So ia, Ana, & González-Aguila , Juan Manuel (2021). F om pola iza ion o ha e: Po ai o he spanish poli ical meme. Social Media and Socie y, 7(4) h p s://doi.o g/10.1177/20563051211062920 Pé ez-Cu iel, Concha, & Limón-Naha o, Pila (2019). Poli ical influence s. A S udy o Donald T ump’s pe sonal b and on Twi e and i s impac on he media and use s. Communica ion & Socie y, 32(1), 57-75. h ps://doi.o g/gwdg Plaza-del-A co, Flo Mi iam, Molina-González, M. Dolo es, U eña-López, L. Al onso, & Ma ín- Valdi ia, M. Te esa (2020). De ec ing Misogyny and Xenophobia in Spanish Twee s Using Language Technologies. ACM T ansac ions on In e ne Technology (TOIT), 20(2), 1-19. h ps://do i.o g/gwh Rollne -Lie n, Gö an (2019). El discu so del odio: una lec u a c í ica de la egulación in e nacional. Re is a Española de De echo Cons i ucional, 115, 81-109. h ps://doi.o g/gwdh Twi e (2021). Polí ica ela i a a las conduc as de inci ación al odio. Twi e . h ps://bi .ly/3w03uY3 Valdez-Apolo, Ma ía Belén, A cila-Calde ón, Ca los, & Jiménez-Amo es, Ja ie (2019). El discu so del odio hacia mig an es y e ugiados a a és del ono y los ma cos de los mensajes en Twi e . RAEIC, Re is a de la Asociación Española de In es igación de la Comunicación, 6(12), 361-384. h ps://doi.o g/dxj2 Williams, Ma hew L., Bu nap, Pe e, Ja ed, Ami , Liu, H.an,& Ozalp, Se a (2020). Ha e in he Machine: An i-Black and An i-Muslim Social Media Pos s as P edic o s o Offline Racially and Religiously Agg a a ed C ime. The B i ish Jou nal o C iminology, 60(1), 93-117. h ps://doi.o g/c9qh Zamo a Medina, Rocío, Ga ido Clemen e, Pila & Sánchez Ma ínez, Jo ge (2021). Analysis o ha e speech in ol ing islamophobia on wi e and i s social epe cussion in he case o he campaign “Remo e he labels om he eil”. Anàlisi: Quade ns de Comunicació i Cul u a, 65, 1-19. h ps://doi.o g/10.5565/REV/ANALISI.3383