Automated Essay Scoring Using the KNN Algorithm

Author

Bin, Li ; Jun, Lu ; Jian-Min, Yao ; Qiao-Ming, Zhu

Author_Institution

Provincial Key Lab. of Comput. Inf. Process. Technol., Soochow Univ., Suzhou

Volume

1

fYear

2008

fDate

12-14 Dec. 2008

Firstpage

735

Lastpage

738

Abstract

The K-Nearest Neighbor (KNN) algorithm for text categorization is applied to CET4 essays. In this paper, each essay is represented by the vector space model (VSM). After removing the stop words, we chose the words, phrases and arguments as features of the essays, and the value of each vector is expressed by the term frequency and inversed document frequency (TF-IDF) weight. The TF and information fain (IG) methods are used to select features by predetermined thresholds. We calculated the similarity of essays with cosine in the KNN algorithm. Experiments on CET4 essays in the Chinese Learner English Corpus (CLEC) show accuracy above 76% is achieved.

Keywords

educational administrative data processing; pattern classification; text analysis; CET4 essays; CLEC; Chinese Learner English Corpus; KNN algorithm; VSM; automated essay scoring; inversed document frequency weight; k-nearest neighbor algorithm; term frequency weight; text categorization; vector space model; Computer science; Frequency; Laboratories; Machine learning; Performance gain; Software algorithms; Software engineering; Space technology; Text categorization; Writing; KNN algorithm; automatic essay scoring; feature selection; vector space model;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Software Engineering, 2008 International Conference on

Conference_Location

Wuhan, Hubei

Print_ISBN

978-0-7695-3336-0

Type

conf

DOI

10.1109/CSSE.2008.623

Filename

4721854