Siv Tau Ntev Ntev Ntev Memory-based Sentiment Analysis Of E-Commerce Reviews Part 3

4. Cov txiaj ntsig

Peb tau cob qhia peb tus qauv rau li 10 lub sijhawm thiab suav cov kev cob qhia thiab kev siv tsis tau zoo nrog rau kev cob qhia thiab kev siv tau.

Kev poob yuav zoo li ib qho xwm txheej uas tsis tuaj yeem nyob hauv peb lub neej, tab sis nws qhov cuam tshuam rau peb lub hlwb thiab kev nco tuaj yeem ua rau muaj txiaj ntsig.

Ua ntej, thaum peb ntsib kev poob, nws yooj yim kom tau txais kev xav tsis zoo. Qhov no yuav suav nrog kev tu siab, npau taws, ntxhov siab, thiab lwm yam. Cov kev xav no tuaj yeem cuam tshuam rau peb lub hlwb, ua rau peb qaug zog, tsis muaj zog, thiab poob siab. Nyob rau tib lub sijhawm, cov kev xav tsis zoo no kuj tuaj yeem cuam tshuam rau peb lub cim xeeb, ua rau nws nyuaj rau peb mloog thiab peb lub cim xeeb tsis meej.

Nyem paub ntxiv los txhim kho kev nco

Txawm li cas los xij, txawm tias thaum peb poob, peb tseem tuaj yeem tuav tus cwj pwm zoo thiab kev xav zoo, pab peb kov yeej qhov kev xav tsis zoo ntawm peb lub hlwb thiab kev nco. Qhov no suav nrog kev siv sijhawm nrog phooj ywg thiab tsev neeg, sib qhia kev xav, tswj kev noj qab haus huv thiab kev ua neej nyob, koom nrog kev tawm dag zog lub cev, thiab koom nrog kev kho mob hlwb, thiab lwm yam.

Tsis tas li ntawd, peb tuaj yeem ua cov kauj ruam tseem ceeb los pab peb tus kheej kom muaj kev nco zoo thaum lub sijhawm poob. Qhov no suav nrog muab sijhawm rau koj tus kheej los lees txais thiab hloov kho rau cov xwm txheej tshiab, nrog rau kev sim cov tswv yim thiab cov tswv yim tshiab, xws li kev tsim cov koom haum nco, rov ua dua, siv cov cim nco, thiab lwm yam.

Tag nrho hauv tag nrho, kev poob ua rau peb lub hlwb kev noj qab haus huv thiab kev nco, tab sis peb tuaj yeem ua cov kauj ruam zoo los kov yeej cov teebmeem no. Yog tias peb muaj peev xwm tswj tau tus cwj pwm zoo thiab muab sijhawm rau peb tus kheej los hloov kho cov xwm txheej tshiab, peb yuav tuaj yeem rov qab tau yooj yim dua peb txoj kev noj qab haus huv thiab kev nco thiab txav mus rau lub neej zoo dua. Nws tuaj yeem pom tau tias peb yuav tsum txhim kho kev nco, thiab Cistanche deserticola tuaj yeem txhim kho kev nco zoo vim Cistanche deserticola yog cov khoom siv tshuaj suav tshuaj uas muaj ntau yam teebmeem, ib qho ntawm kev txhim kho kev nco. Kev ua tau zoo ntawm cov nqaij minced los ntawm ntau yam khoom xyaw uas nws muaj, nrog rau cov kua qaub, polysaccharides, flavonoids, thiab lwm yam. Cov khoom xyaw no tuaj yeem txhawb lub hlwb kev noj qab haus huv ntau txoj hauv kev.

Peb tuaj yeem pom los ntawm daim duab 4 tias ob qho tib si kev cob qhia thiab kev siv tau raug txo qis hauv kev cob qhia ntawm tus qauv. Daim duab 5 qhia tau hais tias kev cob qhia thiab kev siv tau qhov tseeb tau nce ntxiv tom qab rau 10 lub sijhawm.

Txij li thaum, tom qab kev twv ua ntej, cov txiaj ntsig kawg peb tau txais yog qhov tsim nyog, peb siv qee qhov pib los txiav txim siab seb cov ntaub ntawv puas yog cov chav kawm zoo lossis tsis zoo. Rau lub hom phiaj no, peb tau siv ROC nkhaus uas npaj cov nqi tseeb thiab qhov tsis zoo.

Nws pab nrhiav qhov muaj nuj nqis rau qhov binary classifier. Los ntawm peb lub ROC curveshown nyob rau hauv daim duab 6, peb tau xaiv 0.78 raws li peb theem.

+e cov qauv khaws tseg yog rov qab ua dua thiab kev kwv yees tau tsim tawm ntawm cov ntaub ntawv xeem uas xav txog tus nqi saum toj no. Tam sim no peb muaj qhov kev xav zoo li qub thiab qhov kev xav tau kwv yees.

Txij li cov dataset tsis sib npaug, qhov zoo dua qhov ntsuas tus qauv yuav yog F1 tus qhab nia es tsis yog qhov tseeb.

Hauv Table 4, peb tau sau qhov tseeb, qhov tseeb, rov qab, thiab F1 cov qhab nia ntawm lwm cov qauv hauv qab thiab muab piv nrog peb cov qauv. +e cov qauv hauv paus tau txiav txim siab los ntawm cov ntaub ntawv peb tau tshuaj xyuas rau qhov kev sim no.

5. Xaus + yog daim ntawv sib tham txog kev soj ntsuam kev xav hauv cov ntsiab lus ntawm kev lag luam tshuaj xyuas. Muaj ntau yam kev tshawb fawb yav dhau los hauv thaj tsam ntawm kev xav mining ntawm kev tshuaj xyuas.

Peb cov ntaub ntawv muaj cov tshuaj xyuas los ntawm cov xov tooj ntawm tes thiab cov khoom siv ntawm Amazon. Long Short-TermMemory Networks tau siv los faib cov kev xav siv kev kawm tob. Peb cov ntaub ntawv qhia kev cai tau siv los rho tawm cov yam ntxwv kos hauv word2vec embeddingtechnique. Raws li ROC nkhaus, peb txiav txim siab tias 0.78 yog qhov kawg uas peb yuav tsum siv los faib cov kev xav.

Plaub qhov ntsuas tau siv los ntsuas peb tus qauv kev ua tau zoo: qhov tseeb, qhov tseeb, rov qab, thiab F1 tau qhab nia. Aprecision ntawm 97% pom tias yog qhov siab tshaj plaws ntawm plaub parameters.

Raws li cov ntaub ntawv tsis sib npaug, peb txiav txim siab F1score yog qhov kev ntsuas zoo tshaj plaws ntawm tus qauv kev ua tau zoo, uas ua rau kev ntsuam xyuas ntawm 93%. Qhov kev sim tseem ceeb ntawm qhov kev tshawb fawb no yog los ntsuas kev ua haujlwm ntawm tus qauv nrog ntau cov ntaub ntawv. + yog txoj kev muab cov txiaj ntsig zoo txawm tias cov ntaub ntawv loj li ntawm 938,261 kev tshuaj xyuas. +e Qhov tseem ceeb ntawm kev siv cov qauv no yog tias LSTM yuav tsum xav txog kev nco mus ntev thiab word2vec muaj txiaj ntsig zoo ntawm cov lus sawv cev uas pab hauv kev txheeb xyuas cov txiaj ntsig zoo.

Rau kev ua haujlwm yav tom ntej, peb xav xav txog kev siv bidirectional LSTM rau kev faib tawm kev xav uas cob qhia ob txoj hlua ntawm LSTMs, qhov tseeb inputsequence thiab rov qab ib. + yog yuav pab txhim kho cov qauv kev ua tau zoo.

Cov ntaub ntawv muaj

+e cov ntaub ntawv siv los txhawb qhov kev tshawb pom ntawm qhov kev tshawb fawb no muaj los ntawm tus sau sau raws li kev thov.

Kev tsis sib haum xeeb

+e sau ntawv tshaj tawm tias lawv tsis muaj kev tsis sib haum xeeb ntawm kev txaus siab.

Kev lees paub

+ yog kev tshawb fawb tau nyiaj txiag los ntawm Ntxhais fuabtais Nourah bint Abdulrahman University Cov Kws Tshawb Fawb Txhawb Txoj Haujlwm tus lej.(PNURSP2022R120), Ntxhais fuabtais Nourah bint AbdulrahmanUniversity, Riyadh, Saudi Arabia.

Cov ntaub ntawv

[1] B. Liu, Sentiment Analysis: Mining Opinions, Sentiments, and Emotions, Cambridge Univ. Xovxwm, New York, NY, USA, 1stedition, 2015.

[2] P. Balaji, O. Nagaraju, thiab D. Haritha, "Lub Peb Hlis). 436–439, IEEE, Chirala, Is Nrias teb, Lub Peb Hlis 2017.

[3] R. Varghese thiab M. Jayasree, "Ib daim ntawv ntsuam xyuas ntawm kev xav thiab kev xav mining," International Journal of RenewableEnergy Technology, vol. 2, nre. 11, pp. 312–317, 2013.

[4] C. Sindhu thiab DS Chandrakala, "Kev tshawb fawb ntawm kev xav txog kev xav thiab kev xav polarity kev faib tawm," InternationalJournal of Emerging Technology thiab Advanced Engineering, vol. 3, pp. 531–539, 2013.

[5] A. Jurek, Mulvenna, thiab Y. Bi, "Kev txhim kho lexicon-raws li kev tsom xam rau kev sib raug zoo xov xwm analytics," Security Informatics, vol. 4, nre. 1 ib,p. Peb Hlis 9, 2015.

[6] . 8, pp. 189513–189526, 2020.

[7] J. Devlin, M. Chang, K. Lee, thiab K. Toutanova, "BERT: pretraining of deep bidirectional transformers for languageunderstanding," in Proceedings of the Conference NorthAmerican Chapter Association Computational Linguistics, Human Language Technol, pp. 4171 -4186, Minneapolis, Minnesota, Lub Rau Hli 2019.

[8] L. Dey, S. Chakraborty, A. Biswas, B. Bose, thiab S. Tiwari, "Sentment analysis of review datasets using naive Bayes andk-nn classifier," 2016,https://arxiv.org/abs/1610.09982.

[9] MR Huq, A. Ali, thiab A. Rahman, "Sentment analysis onTwitter data using KNN and SVM," International Journal of Advanced Computer Science and Applications, vol. 8, nre. 6 ,pab. 19-25, 2017.

[10] BS Lakshmi, PS Raj, thiab RR Vikram, "Sentment analysisusing deep learning technique CNN with KMeans," International Journal of Pure and Applied Mathematics, vol. 114, Nr. 11, p. 47–57, 2017.

[11] Y. Fang, H. Tan, and J. Zhang, "Multi-strategy sentimentanalysis of consumer reviews based on semantic fuzziness,"IEEE Access, vol. 6, pp. 20625–20631, 2018.

[12] B. Shin, T. Lee, thiab JD Choi, "Lexicon integrated CNNmodels nrog rau kev saib xyuas kev xav," 2016,https://arxiv.org/abs/1610.06272.

For more information:1950477648nn@gmail.com