| Titre : | Application of Deep Learning in Sentiment Analysis | | Type de document : | projet fin études | | Auteurs : | Sanaa Lakrouni, Auteur | | Langues : | Français (fre) | | Catégories : | SDBD
| | Index. décimale : | mast 96/18 | | Résumé : | This paperaddressestheproblemofsentence-levelsentimentanalysiswithdeep
learning system,wehavebeenseenthousandsofpaperthatproposedmultiplema-
chinelearningmethodssomeofthemattainedgreatresultsbutinrecentyears,with
the succefulresultsthatdeeplearningshowsindifferentfieldwiththeirdifferent
architectureConvolutionandRecursiveNeuralNetworkshavebeenproventobe
effectivenetworkarchitectureforsentencelevelsentimentanalysis,inthisstudy
wewillintroducethedifferentreasontousedeeplearningwithsentimentanalysis
and thedetailsthathavebeenaddedtocombinedeeplearningandsentimentana-
lysis ,wedescribeitsdifferentarchitectures,takingadvantageofthewaythatdeep
learning generateautomaticallyefficientfeaturesandusethemintoclassification
to attainedsufficientresults.IntheothersecondExperiencewithtweetdatashow
the usualresultswithCnnandlstmandacombiningarchitecturebetweenthose
architecturehavebeentrainedwithalocalmachine.usingtweetsasourdatawas
the objectivepurposeofthisstudytoadaptthisfromEnglishlanguageintothe
dialect Moroccans.InthiscontextwetriedtodoSentimentanalysiswithmoroccan
dialect usingthesameprocessthatwehavebeendonebeforewithenglishdataand
compare results. |
Application of Deep Learning in Sentiment Analysis [projet fin études] / Sanaa Lakrouni, Auteur . - [s.d.]. Langues : Français ( fre) | Catégories : | SDBD
| | Index. décimale : | mast 96/18 | | Résumé : | This paperaddressestheproblemofsentence-levelsentimentanalysiswithdeep
learning system,wehavebeenseenthousandsofpaperthatproposedmultiplema-
chinelearningmethodssomeofthemattainedgreatresultsbutinrecentyears,with
the succefulresultsthatdeeplearningshowsindifferentfieldwiththeirdifferent
architectureConvolutionandRecursiveNeuralNetworkshavebeenproventobe
effectivenetworkarchitectureforsentencelevelsentimentanalysis,inthisstudy
wewillintroducethedifferentreasontousedeeplearningwithsentimentanalysis
and thedetailsthathavebeenaddedtocombinedeeplearningandsentimentana-
lysis ,wedescribeitsdifferentarchitectures,takingadvantageofthewaythatdeep
learning generateautomaticallyefficientfeaturesandusethemintoclassification
to attainedsufficientresults.IntheothersecondExperiencewithtweetdatashow
the usualresultswithCnnandlstmandacombiningarchitecturebetweenthose
architecturehavebeentrainedwithalocalmachine.usingtweetsasourdatawas
the objectivepurposeofthisstudytoadaptthisfromEnglishlanguageintothe
dialect Moroccans.InthiscontextwetriedtodoSentimentanalysiswithmoroccan
dialect usingthesameprocessthatwehavebeendonebeforewithenglishdataand
compare results. |
|
Faire une suggestion Affiner la recherche
Application of Deep Learning in Sentiment Analysis / Sanaa Lakrouni
