This study reviews the developments of automated scoring systems and techniques over the past 40 years and the strengths and weaknesses of each technique are analyzed in detail. Results of the study showed that automated scoring systems are becoming more complicated and advanced by the integrated use of knowledge in natural language processing, statistics, information retrieval, corpus linguistics, etc. but they still have drawbacks in one or two aspects. The study provides some valuable insights into further studies in this line. PDFFulltextXMLReferencesCitation
How to cite this article
Jinlin Jiang and Wei Wei, 2012. Automated Scoring Research over 40 Years: Looking Back and Ahead. Journal of Artificial Intelligence, 5: 56-63.