Study-Chinese

Nazarin hulɗar malami-ChatGPT wajen tsara ayyukan sadarwa masu dogaro da rikici don kwasa-kwasan magana na matakin jami'a, tare da bincika rawar da tasirin AI.

Exploring the use of Answer Set Programming to formalize VanPatten's Input Processing Theory, enabling automated prediction of learner interpretations to assist language teaching.

Yana binciken fassarar ma'ana ta Sinanci zuwa DRS ba tare da bayanan da aka yiwa lakabi ba, yana ba da shawarar tsarin tattara bayanai da gwajin gwaji don kimantawa mai zurfi, yana nuna kalubale tare da maganganu.

Gabatar da CPG-EVAL, ma'auni na farko na tsari don tantance ilimin manyan harsunan AI game da nahawun koyarwa a cikin yanayin koyar da Sinanci.

Bincike kan tsarin gyara kurakuran rubutu na Sinanci, yana nuna yadda tsarin kuskure ya yi yawa fiye da kima yayin da tsarin harshe a cikin BERT bai isa ba, tare da shawarar dabarar rufe bazuwar don ingantaccen fahimta.

Wata sabuwar hanya ta Gyaran Kuskuren Rubutun Sinanci (CSC) wacce ke sake tsara jimloli gaba ɗaya maimakon yiwa haruffa alama, ta sami sakamako mafi kyau da ingantaccen amfani.

Ya gabatar da CPG-EVAL, ma'auni na farko don kimanta ilimin nahawun koyarwar Sinanci na Manyan Harsunan Na'ura (LLMs) bisa tsari, yana tantance gane, bambancewa, da juriya ga ruɗani.

Nazari mai zurfi na amfani da Tsarin Ma'ana Mai Zurfi (DeepFM) a cikin aikin Duolingo SLAM don binciken ilimi a matakin kalma.

Yana binciken aikace-aikacen kayan aikin Unix DIFF don ayyukan NLP kamar gano bambance-bambance, ciro dokoki, haɗa bayanai, da daidaitawa mafi kyau, yana nuna amfaninsa da yawan amfaninsa.

Bincike kan sabon tsarin haɗin kai wanda ya haɗa Bishiyoyi na yanke shawara masu Haɓakawa da RNNs don hasashen gibin ilimin ɗalibi a cikin koyon harshe, wanda ya sami maki mafi girma a Aikin Raba na SLAM na 2018.

Nazarin adalci a cikin tsarin hasashen koyon harshe na biyu, tare da kimanta karkata ta kan dandamali da ƙasashe ta amfani da bayanan Duolingo.

HSK study materials for Study-Chinese

Bayani game da Aikin MOSLA, bayanai na musamman na tsawon lokaci, tsaka-tsaki, da harsuna da yawa wanda ke ɗaukar cikakkiyar tsarin koyon harshe na biyu sama da shekaru biyu.

Binciken tsarin KOSHIK don sarrafa Harshe ta Halitta (NLP) mai girma ta amfani da Hadoop, ya ƙunshi sassansa, matakan aiwa, kimanta aiki, da alkiblar gaba.

Wannan takarda tana bincika tsarin KOSHIK don sarrafa harshe na halitta mai girma ta amfani da Hadoop, tana bayyana aiwatar da shi, kimanta aikin sa, da kuma hanyoyin gaba.

Nazarin yanayin NLP (2021-2023) da yuwuwar aikace-aikacen sa wajen inganta sadarwar yawon shakuwa, ciki har da fassarar atomatik da chatbots na AI.

Nazarin yadda ake amfani da takamaiman umarni tare da Manyan Harsunan AI don kaiwa ga matakan CEFR da EBCL (A1, A1+, A2) don koyon Sinanci na musamman.

Takarda bincike tana nazarin canja wurin harshe a cikin samfuran harshe ta amfani da tsarin BERT da bayanan magana ga yara don kwaikwayon Koyon Harshe na Biyu.

Nazarin yadda tsarin harshe na jijiyoyi (LMs) ke koyon harshe na biyu (L2), tare da bincika tasirin horar da harshe na farko (L1), tsarin canja harshe, da kuma fahimtar harshe.

Ya gabatar da sabon tsari na koyar da Sinanci a matsayin harshen waje, wanda ya haɗa Rongzhixue, AI, da tsarin malam-buda na fassara kafin fassara don horar da tunani na harshe biyu.

Bincike na zahiri wanda ya kwatanta amfani da fi'ili da suna a jaridu na Sinanci da na Turanci, da tasirinsa ga rubuce-rubucen masu koyon Sinanci masu magana da Turanci.

Nazarin takarda da ta bincika tasirin kwaikwayon Duniyar Gaskiya ta Kwaikwayo akan ƙarfafawa a cikin koyon harsunan waje, gami da hanyoyin bincike, sakamako, da abubuwan da za a iya gani a nan gaba.