Open AccessA Combination of Text Similarity And Latent Semantic Analysis (LSA) Methods In Automatic Scoring Essay
Author(s) -
Deddy Atmajaya,
Purnawansyah,
Siti Rahayu
Publication year - 2019
Publication title -
international journal of engineering and advanced technology
Language(s) - English
Resource type - Journals
ISSN - 2249-8958
DOI - 10.35940/ijeat.e1206.0585c19
Subject(s) - latent semantic analysis , similarity (geometry) , test (biology) , computer science , reading (process) , semantic similarity , natural language processing , information retrieval , key (lock) , artificial intelligence , mathematics education , psychology , linguistics , paleontology , philosophy , computer security , image (mathematics) , biology
Description of type exam (essay) is considered by many experts as the most appropriate test to reap the results of a complex learning activities, because essay writing will involve the student’s ability to remember, organize, express, and integrate the ideas of the students. Just to correct the essay exam results, requiring a longer time if done manually because most do by reading an essay one by one. So that, lecturers needs to spend a lot of time to assess the answers of student’s exam. Therefore, in implementation, automatic scoring system is needed on the answer essay exam. Automated essay assessment method used in this study is a combination of Text Similarity and Latent Semantic Analysis (LSA) to look for a match and similarity level student answers with the answer key that has previously been inputted into the system. Data used in this test is 15 students with each student to answer 5 questions. Data obtained from subjects essay Basis Data I. The correlation results of that two assessment shows grades 0,946085 with an average increment of 2,08. Which means the results of the assessment system is not much different from the results of the assessment of the lectures, so that the automatic scoring system can be applied to essay type exam.