View
2
Download
0
Category
Preview:
Citation preview
Title Learning false friends across contexts
Author(s) Abou-Khalil, Victoria; Brendan; Flanagan; Ogata, Hiroaki
Citation Companion Proceedings 8th International Conference onLearning Analytics & Knowledge (LAK18) (2018)
Issue Date 2018
URL http://hdl.handle.net/2433/233070
Right
This work is published under the terms of the CreativeCommons Attribution- Noncommercial-ShareAlike 3.0Australia Licence. Under this Licence you are free to: Share -copy and redistribute the material in any medium or format.
Type Conference Paper
Textversion publisher
Kyoto University
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
1
Learning false friends across contexts
Author(s):VictoriaAbou-KhalilGraduateSchoolofInformatics,KyotoUniversity
v.aboukhalil@gmail.com
Author(s):BrendanFlanaganAcademicCenterforComputingandMediastudies,KyotoUniversity
flanagan.brendanjohn.4n@kyoto-u.ac.jp
Author(s):HiroakiOgataAcademicCenterforComputingandMediastudies,KyotoUniversity
hiroaki.ogata@gmail.com
ABSTRACT: False friends arewords in two languages that look or sound similar but differsignificantly in meaning in some or all contexts. False friends are confusing for languagestudentsandcouldresultinfrustrationandcommunicationproblems.Thispaperproposesamethod to diagnose and prevent false friends mistakes based on students’ past learnedwords, current location and time. The proposed method uses records from the SCROLLsystem (System for Capturing and Reminding Of Learning Log) to analyze the previousactivityofstudents.Weassumethatthepastactivityofastudentcanbeusedtopredictthemeaningintendedbythestudentwhenlookingupapolysemousword.Theidentificationofthe intendedmeaning in thestudent'scurrentcontext is thenusedtoprovidethestudentwith the appropriate translation, warnings and quizzes, possibly improving the learningprocessandavoidingfalsefriendsfuturemistakes.
Keywords: Learning Analytics, Ubiquitous learning, False Friends, Computer SupportedLanguageLearning
1 INTRODUCTION
When learning a second language, students can take advantage of the vocabulary of their firstlanguageusingcognates (Nation,2003).Cognatesarewords that soundor looksimilar in the twolanguages,havesimilarmeanings,andhelpstudentsexpandtheirvocabularybyplayingtheroleof`truefriends'.
However,wecanallimaginetheawkwardsituationsthatcouldariseiftheFrenchwordpromiscuité(lack of privacy, crowding) is interchangedwith the Englishwordpromiscuity (i.e., having a lot ofdifferent sexual partners) in a sentence.We can also imagine flirting goingwrongwhen a Frenchspeaker compliment their English crush using the word formidable (i.e., inspiring fear or respectthrough being impressively large or powerful), when what she actuallymeant is formidable (i.e.,inspiring awe).We can understand how strange a Japanese speaker will soundwhen saying `Mymansion is on the second floor',whenwhat he had inmindwasマンション (mansion, i.e., flat,
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
2
apartment). The previous situations are faced by language learners and are caused by a trickycategoryofwords:falsefriends.
False friends are words in two languages that look and sound similar, but differ significantly inmeaning in some or all contexts. The degree of complexity of learning false friends depends onwhether they are total false friends or partial false friends. Total false friends have completelydifferentmeanings in both languages (e.g.: Eng.:Attend (to be present); French:Attendre (wait)),whereaspartial falsefriendsarepolysemouswords,oneofwhosemeanings isafalsefriendwhileothersaretruecognates(e.g.:Eng.:Demand(i.e.,requestmadeasofright);French:Demander(i.e.,to ask; to be looking for; to demand). Figure 1 shows the types of `friends' that a student willencounterwhenlearningthetargetlanguage.
Figure1:Typesoffalsefriendsbetweenastudent'snativelanguageandthelearnedtargetlanguage.
Dependingon thecontext,partial false friendsplay the roleofa trueora false friend.Thispaperproposesamethodtodiagnoseandpreventfalsefriendsmistakesbasedonstudents'pastlearnedvocabulary, current location and time. First, we analyze the factors that lead to false friends
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
3
confusionswhile learning the target language.Wealsoanalyze the typesofassimilationproblemsthat could then arise, depending on the type of false friend. In order to prevent false friendsconfusions,weuse the student'spast learning logs (previous lookedupwords, locations, time) topredict the meaning intended by the student when looking up a polysemous words. Theidentification of the intendedmeaning is then used to provide the student with the appropriatetranslationandwarnings,possibly improving the learningprocessandavoiding futuremistakes. Inorder to fortify the learning the student knowledge will be tested. The language learner will getquizzesaboutthemeaningoflearnedpolysemouswordsindifferentcontexts(Location,time).
2 DIFFICULTIES AND PROBLEMS OF FALSE FRIENDS LEARNING
From the pedagogical perspective, intrinsic and extrinsic factors determine the degree of falsefriends'difficultyfor language(Beltran,2006).Theintrinsicfactorscontributingtoahigherlevelofdifficultyoffalsefriendslearningare:
IF1:Theconfusingnatureoffalsefriends,andparticularlythepartialfalsefriends.Somefalsefriendshave always a deceptive meaning, whereas some others have deceptive meanings in certaincontexts only. This creates uncertainty for students as they could fail to recognize in whichcontextsthewordisafalsefriend,andinwhichcontextsitisnot.
IF2:Semanticfieldsmayoverlap.Wordscanhavedifferentmeaningsinbothlanguages,butbelongtothesamesemanticfield(e.g.:Japanese:フィルム.firumumeanscamerafilmroll).
IF3: Because of the large number of true friends, students have a tendency to overgeneralize thewordsthattheycomeacross.
Theextrinsicfactorscontributingtothecomplexityoffalsefriendslearningare:
EF1: Language learners are usually encouraged to take advantage of true cognates without beingwarnedof theexistenceof false friends.This could lead to frustration for the language learnerwhentheynoticetheactualcomplexityofthecognates.
EF2: Oversimplification of dictionaries by lexicographers where translations sometimes lack ofnuancesandcontexts.
Depending on the type of false friend, the previous factors influence differently the difficulty inlearningthem.Inordertoimprovefalsefriendsteaching,itisimportanttounderstandwhichtypeoffactorsinfluencefalsefriendslearning.Table1showswhichfactorsaffectwhichtypeoffalsefriendslearningaswellastheoveralllearningdifficulty.Xisdisplayedwhenatypeoffalsefriendslearningisnotaffectedbyafactor.Oisdisplayedwhenatypeoffalsefriendisaffectedbyafactor.
Table1:FactorsaffectingFalseFriendslearningandlearningdifficulty.
F2:FalseFriends
F3:PartialFalseFriendsN∩T
F4:PartialFalseFriendsT⊂N
F5:PartialFalseFriendsN⊂T
IF1 X O O X
IF2 O X X X
IF3 O O O X
EF1 O O O O
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
4
EF2 X O O O
Difficulty Medium High High Low
Falsefriendsarerarelyincorporatedintolanguageclassesdespitethedifficultiesfacedbylanguagestudentswhendealingwiththem.Whentheyarepointedoutbytheteachers,thewords’nuancesareoftenover-simplifiedanddowngradetheaccuracyandassimilationofthemeaningoftheword(Hayward,1984). Inthecaseofpartial falsefriends,this lackofaccuracycan leadtotwodifferentsituations:
• alossofsomemeaningsofthewordinthetargetlanguage.• anadditionofsomemeaningstotheword in thetarget languagebyprojectingthemeaningof
thewordinthenativelanguagetothewordinthetargetlanguage.
Figure2showsthesituationswheretheoversimplificationleadstothelossoradditionofmeanings.
Figure2:Lossorprojectionofmeaningsduetooversimplifiedteachingoffalsefriends,dependingonthetypeoffalsefriend.
Inordertoavoidthiskindofmistakes,falsefriendsneedtobepointedoutbytheteachersandthecorrect translations should be given (August, 2005). Moreover, the meanings should not be
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
5
presentedasonebloc to thestudentbutshouldbeput inacontextandrestricted toaparticulararea(Hayward,1984).
However,inaselflearningenvironment,teachershavelessornocontroloverthelearnedcontent.With the rise of smart phones, the use ofmobile devices in language learning is a growing trend(Godwin, 2016). Student have more control over their learning pace (Benson, 2005) and thevocabularytheywanttolearn,butdon'thaveopportunitiestocommunicatewhichmeaningofthewordtheyarelookingfor.
Thispaperproposestotakeadvantageofstudentpast learnedwordstounderstandtheparticularmeaning queried by the learner and provide themwith the correct translation in their intendedcontext.Moreover,we aim atminimizing the effect of the intrinsic and extrinsic factors affectingfalsefriendslearningbyshowingthelearnerthedifferentmeaningsandnuancesofthewords.Theassimilationofthestudentwillbethentestedinthecontextofusageofthewordinordertofortifytheirknowledge.
3 METHOD
3.1 SCROLL system
DuringthisstudyweuserecordsfromtheSCROLLSystem(SystemforCapturingandRemindingOfLearningLog).Scrollisadigitalrecordofwhatlanguagestudentshavelearnedindailylife.Itallowsthelearnerstologthenewwordsorsentencestheylearnedalongwithphotos,audios,videosandlocation (Ogata, 2011). SCROLL captureswhat learners are learning aswell as its contextual data.Theusersarethenremindedofwhattheylearnedintherightplaceandtherighttime.Moreover,students receive personalized quizzes to fortify the learning. Figure 3 is a screenshot from theSCROLLsystemthatshowsaloginsertedbyastudentforthewordKaraoke.Thestudentappendeda picture and a location when creating the log. A Japanese translation of the word Karaoke isautomaticallyprovidedtothestudent,andthetimeisautomaticallyregistered.
Figure3:ScreenshotfromtheSCROLLsystemshowingaloginsertedbyastudent.
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
6
Thelogswewillbeusingincludemeta-datasuchas:
Knowledge:wordsthatstudentshavelearnedinthepastUser:authoridentificationPlace:placewherethelearninghappened(cinema,restaurant)Time:timewhenthelearningtookplace
Currently SCROLLhas1648usersand contains24355 logs. The system isusedmainlyby studentslearningJapanese.
3.2 Contextual false friends learning
WhenusingtheSCROLLsystem,JapaneselanguagelearnersinsertlogscontainingawordinEnglishandlearnitsJapanesetranslation.However,ifthewordisafalsefriend,studentsgetatranslationthatdoesnotusually reflect the context, thedifferentmeaningsand thenuancesof theword. Inordertoprovidelearnerswiththerighttranslationintherightcontext,wehavetounderstandtheirintendedmeaning.Weproposetoanalyzethepastactivityofstudentstoextractthemeaningtheyare looking for.Weassumethat thepastactivityofastudentgivesacontext thatcanbeusedtoextract themeaning intendedby the studentwhen lookingup a polysemousword. The followingmeta-dataareusedtopredictthemeaningthestudentislookingfor:
Pastknowledge(words): tomeasurethesimilaritybetweenthecurrentwordthestudent is lookingupandpastwordsthestudentlookedupinthesystem.
Time: tomeasure the similaritybetween thecurrentword the student is lookingupand the setofwordsthestudentlookedupatthesameperiodoftime.
Location:tomeasurethesimilaritybetweenthewordthestudentislookingupandthesetofwordsthestudentlookedupatthesamelocation.
WeproposetomeasurethesimilarityusingJiangsimilaritymeasure.JiangsimilaritymeasurebasedinformationcontentofeachconceptinWordNet.Itassumesthateachconceptincludesinformationin WordNet and the more common information two concepts share, the more similar the twoconcepts are (Meng, 2013). Previous studies where semantic distance measures were comparedexperimentallyfoundthatJiang'smeasuregavethebestresultsoverall(Budanitsky,2001).
Figure4showsanexample.Thestudent lookeduptheworddriver.However,theworddriverhasdifferentmeaningsanddifferenttranslationsinJapaneseforeachofthosemeaning:
Personwhodrivesavehicle:ドライバ(doraiba)
Traindriver:運転⼠(untenshi)
Massdriver:マスドライバー(masudoraiba)
Computing:仮想デバイスドライバ(kasoodebaisudoraiba)
Golfclub:ドライバー(doraiba)
Screwdriver(BritishEnglish):ドライバー(doraiba)
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
7
Figure4:ScreenshotfromtheSCROLLsystemshowingthestudentpastknowledgeofastudentlookingforthemeaningoftheworddriver
Inordertounderstandwhichmeaningthestudentislookingfor,welookatthepreviousknowledge,thewordslookedupatthesameperiodoftimethantheworddriverandthewordslookedupatthesame location of theworddriver. In this case, theword looked up just before theworddriver isstepladder.Thesemanticdistancebetweenthewordssrewdriverandstepladderissmallerthanthesemantic distance between the word stepladder and each of the words: car driver, train driver,computerdriver,massdriverandgolfclub.Wecanconcludethatthemeaningthestudentislookingforisscrewdriver,andthatthetranslationthatshouldbegivenisドライバー(doraiba).
3.3 Preventive false friends learning
As stated previously, students face intrinsic and extrinsic factors that make false friend learningmoredifficult.Weproposetotacklethefactorsbyprovidingpreventivefalsefriendslearning.Upontheencounterofanewfalsefriend,thestudentwillreceiveawarningthatshowsacomprehensivelist of translations in different contexts. The displayedwarning is different for each type of falsefriendsasshownintable2.
Table2:Warningfordifferenttypeoffalsefriends..
FalseFriendType Typeofwarningreceived
F2:FalseFriends
Warningstatingthatthewordisafalsefriend.
F3:PartialFalseFriendsN∩T
Warningthat:
• statesthatthewordisapartialfalsefriend.
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
8
• liststhedifferenttranslationsforthedifferentmeaningsoftheword.
• liststhecognatesofthewordinthetargetlanguagethathavemeaningsthatarenon-existentinthenativelanguage.
F4:PartialFalseFriendsT⊂N
Warningthat:
• statesthatthewordisapartialfalsefriend.
• itemliststhedifferenttranslationsforthedifferentmeaningsoftheword.
F5:PartialFalseFriendsN⊂T
Warningthat:
• statesthatthewordisapartialfalsefriend.
• liststhecognatesofthewordinthetargetlanguagethathavemeaningsthatarenon-existentinthenativelanguage.
Theexpectedconsequencesofdisplayingwarningsare:
Consequence1: Awareness that the words are used differently in different contexts: The warningstatesthefalsefriendtype.
Consequence2:Knowledgethatavoidsover-generalization(additionofmeanings):Thewarningliststhedifferenttranslationsindifferentcontexts.
Consequence3: Knowledge that avoids over-simplification (loss ofmeanings): Thewarning lists thedifferenttranslationsindifferentcontexts.
Table 3 shows how the different consequences reduce the effects of the intrinsic and extrinsicfactorsdependingonthetypeoffalsefriends.
Table3:Consequencesaffectingthereductionofintrinsicandextrinsicfactorsdependingonthetypeoffalsefriends.
F2:FalseFriends
F3:PartialFalseFriendsN∩T
F4:PartialFalseFriendsT⊂N
F5:PartialFalseFriendsN⊂T
IF1 X Consequence1 Consequence1 X
IF2 Consequence1 X X X
IF3 Consequence1 Consequence1 Consequence1 X
EF1 Consequence1 Consequence1 Consequence1 Consequence1
EF2 X Consequence2,Consequence3
Consequence2,Consequence3
O
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
9
In the previous example, the student looked up the word driver. After predicting the intendedmeaning of the student, the student will be provided with the word ドライバー (doraiba) as a
translation.However,driverandドライバーarepartialfalsefriendsoftype4(F4).Inthiscase,the
student might use the wordドライバー (doraiba) as they would have used the word driver in
EnglishtoexpressTraindriver,massdriverordriverincomputing.Toavoidthismistake,weproposetodisplayawarningtothestudentshowingthedifferenttranslationsoftheworddriver,dependingon the meaning (as shown in Figure 5). The warning gives the student awareness about thecomplexityoftheword,andconsciousnessthatthetranslationprovidedbythesystemcanbeusedinparticularcontextsonly.
Figure5:Warningdisplayingdifferentmeaningsoftheworddependingonthecontext
3.4 Quizzes across contexts
Inorder todetermine theeffectof contextual translationsandwarningson false friends learning,students will be given quizzes to test their acquired knowledge. SCROLL system offers theopportunity to give quizzes to students depending on their location. When the students will bepresent at a location related to one of the meanings of the word a quiz will appear asking thestudentaboutthetranslationofthewordinthisparticularcontextasshowninfigure6.Thequizzeswill be given to the group of students that received the contextual translation as well as thewarnings.Anothergroupofstudentsthatdidn'tgetthecontextualtranslationandthewarningswillplaytheroleofacontrolgroupandbesubjectedtoquizzesaswell.Theresultswillbecomparedtoidentifytheeffectofthepreviousmethodonfalsefriendslearning.
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
10
Figure6:Quizzesdisplayedtothestudentsdependingontheirlocation
4 DISCUSSION
Weproposedamethodtoprevent false friendsmistakes.Themethod isapplied in thecontextofmobilelearning.Themainfeaturesofmobilelearningareaccessibility,immediacy,interactivityandsituating of instructional activities (Ogata, 2004), benefit students during the learning process.However,mobilelearningdoesnotallowlanguagestudentstosharetheirintendedmeaningwhilelooking up a polysemous word. The first part of our work consists of predicting the meaningintended by the student when they are looking up a polysemous word. In order to do so, theproposedmethoduses records from the SCROLL system (System for Capturing andRemindingOfLearning Log) to analyze the previous activity of students. We assumed that the students' pastlearned words, current location and time can be used to predict the meaning intended by thestudentwhen looking up a polysemousword. The identification of the intendedmeaning is thenusedtoprovidethestudentwiththeappropriatetranslation,basedontheintendedmeaning.Thesecond part proposes to display warnings and quizzes to the students. The warnings explain themeaningof thewordandprovidethestudentwithdifferent translations in thedifferentcontexts.Thequizzesaimatfortifyingtheknowledgeofthestudents,possiblyimprovingthelearningprocess.
This method puts into application the theoretical pedagogical approach of false friends learning.Futureworkwillevaluatetheaccuracyofthepredictionofthestudents'intendedmeaning.Wewillalso evaluate the effects of thewarnings by comparing the learning performance of the students
CompanionProceedings8thInternationalConferenceonLearningAnalytics&Knowledge(LAK18)
CreativeCommonsLicense,Attribution-NonCommercial-NoDerivs3.0Unported(CCBY-NC-ND3.0)
11
beforeandafterreceivingthem.Theimpactofthequizzeswillbeevaluatedbycomparingthefalsefriendsrecallrateofthestudentbeforeandafterthedisplayofthequizzes.
5 CONCLUSION
Thispaperproposes amethod todiagnoseandprevent false friendsmistakesbasedon students’past learning activity. The proposed method uses records from the SCROLL system (System forCapturingandRemindingOfLearningLog)toanalyzethepreviousactivityofstudents.Weassumedthatthestudents'pastlearnedwords,currentlocationandtimecanbeusedtopredictthemeaningintended by the studentwhen looking up a polysemousword. The identification of the intendedmeaningisthenusedtoprovidethestudentwiththeappropriatetranslation,warningsandquizzes,possiblyimprovingthelearningprocessandavoidingfuturemistakes.
Acknowledgments
This research was also partially supported by Grant-in-Aid for Scientific Research (S) from JapanSocietyforthePromotionofScience(24220002,2012-2016)andfromMonbukagakusho(16H06304,2016-2021).
REFERENCES
August,D.,Carlo,M.,Dressler,C.,andSnow,C.(2005).Thecriticalroleofvocabularydevelopmentforenglishlanguagelearners.LearningDisabilitiesResearch&Practice,20(1):50–57.
Beltrán, R. C. (2006). Towards a typological classification of false friends (spanish-english). Resla,19:29–39.Benson, P. (2005). Autonomy and information technology in the educational discourse of the
information.Informationtechnologyandinnovationinlanguageeducation,1:173.Budanitsky, A. and Hirst, G. (2001). Semantic distance in wordnet: An experimental, application-
orientedevaluationoffivemeasures.InWorkshoponWordNetandotherlexicalresources,volume2,pages2–2.
Godwin-Jones,R.(2016).Lookingbackandahead:20yearsoftechnologiesforlanguagelearning.Hayward,T.andMoulin,A.(1984).Falsefriendsinvigorated.InLexeter’83Proceedings,pages190–
198.Kay,G.(1995).Englishloanwordsinjapanese.WorldEnglishes,14(1):67–76.Meng, L., Huang, R., and Gu, J. (2013). A review of semantic similarity measures in wordnet.
InternationalJournalofHybridInformationTechnology,6(1):1–12.Nation, P. (2003). The role of the first language in foreign language learning. Asian EFL Journal,
5(2):1–8.Ogata,H.,Li,M.,Hou,B.,Uosaki,N.,El-Bishouty,M.M.,andYano,Y. (2011).Scroll:Supportingto
share and reuse ubiquitous learning log in the context of language learning. Research &PracticeinTechnologyEnhancedLearning,6(2).
Ogata,H.andYano,Y.(2004).Knowledgeawarenessforacomputerassistedlanguagelearningusinghandhelds. International Journal of Continuing Engineering Education and Life LongLearning,14(4-5):435–449.
Recommended