View : 101 Download: 0

Full metadata record

DC Field Value Language
dc.contributor.author김효준-
dc.date.accessioned2020-02-07T16:30:11Z-
dc.date.available2020-02-07T16:30:11Z-
dc.date.issued2020-
dc.identifier.isbn9789811393402-
dc.identifier.issn1876-1100-
dc.identifier.otherOAK-26409-
dc.identifier.urihttp://dspace.ewha.ac.kr/handle/2015.oak/253334-
dc.description.abstractIn this paper, we designed a review analysis automation system to grasp the credibility of the online education matching platform. Web crawling collects and parses reviews and ratings of educators who are atypical data. We will build an emotional dictionary based on the educational field to grasp online educator credibility using collected review data and SO-PMI. We also propose a method for building large - scale learning data based on the emotion dictionary constructed. We proposed a system that provides more reliable review analysis results by measuring the accuracy of emotion dictionary by using deep learning in constructed learning data and evaluation data. Through this, we intend to help judge the credibility of educator in O2O education matching. © 2020, Springer Nature Singapore Pte Ltd.-
dc.languageEnglish-
dc.publisherSpringer-
dc.subjectDeep learning-
dc.subjectLSTM-
dc.subjectRNN-
dc.subjectSO-PMI-
dc.titleDesign Automation System for Review Analysis Affiliation for Online Educator Reliability Prediction-
dc.typeConference Paper-
dc.relation.volume536 LNEE-
dc.relation.indexSCOPUS-
dc.relation.startpage303-
dc.relation.lastpage309-
dc.relation.journaltitleLecture Notes in Electrical Engineering-
dc.identifier.doi10.1007/978-981-13-9341-9_52-
dc.identifier.scopusid2-s2.0-85076860601-
dc.author.googleLee K.-
dc.author.googleKym H.-
dc.author.googleMoon N.-
dc.date.modifydate20200330144613-
Appears in Collections:
경영대학 > 경영학전공 > Journal papers
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

BROWSE