Skip to main navigation Skip to search Skip to main content

Multimodal Scoring Model for Handwritten Chinese Essay

  • Harbin Institute of Technology

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Essay writing plays a critical role in Chinese language skill teaching. With the smart education becomes a hot topic, the demand for automatic essay scoring (AES) has been emerging among teachers and students. Existing works frequently ignore the impact of the visual modal during the scoring process, such as writing quality in terms of neatness or legibility. This paper addresses the problem with a visual-textual integrating perspective and proposes a deep learning based multi-modal AES. Specifically, implicit alignment algorithm is presented to cohere the distinct visual modal and text modal. Methods are tested on a large-scale dataset consisting of over 4000 essays including HSK publicly available samples. The results show that multi-modal AES reduce the MAE of scoring from 1.13 to 1.06, and the implicit alignment algorithm reduces it further to 1.01.

Original languageEnglish
Title of host publicationDocument Analysis and Recognition – ICDAR 2023 - 17th International Conference, Proceedings
EditorsGernot A. Fink, Rajiv Jain, Koichi Kise, Richard Zanibbi
PublisherSpringer Science and Business Media Deutschland GmbH
Pages505-519
Number of pages15
ISBN (Print)9783031416750
DOIs
StatePublished - 2023
Event2023 International Workshops co-located with the 17th International Conference on Document Analysis and Recognition, ICDAR 2023 - San José, United States
Duration: 24 Aug 202326 Aug 2023

Publication series

NameLecture Notes in Computer Science
Volume14187 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2023 International Workshops co-located with the 17th International Conference on Document Analysis and Recognition, ICDAR 2023
Country/TerritoryUnited States
CitySan José
Period24/08/2326/08/23

Keywords

  • Automated Essay Scoring
  • Implicit Alignment
  • Multi-modal Learning

Fingerprint

Dive into the research topics of 'Multimodal Scoring Model for Handwritten Chinese Essay'. Together they form a unique fingerprint.

Cite this