|
|
2014 » Papers » Volume 1 » Unsupervised System for Automatic Grading of Bachelor and Master Thesis 1. UNSUPERVISED SYSTEM FOR AUTOMATIC GRADING OF BACHELOR AND MASTER THESIS Authors: Mosallam Yusra , Rebedea Traian, Chiru Costin, Toma Lukas, Adhana Mulu Weldegebreal Volume 1 | DOI: 10.12753/2066-026X-14-023 | Pages: 165-171 | Download PDF | Abstract
This paper presents an automatic system for evaluation of Bachelor and Master thesis of Computer Science students. In order to be able to fulfill this task, we have used text complexity measures along with other factors to evaluate the students' thesis. Text complexity has been previously used to predict the students' grade level for which to assign a specific reading passage. Also, it has been used in evaluating student's writing in English language classes. However, up to our knowledge it has not been used before to evaluate scientific reports. The main challenges of this task are to select the best features that accurately reflect student's performance, and to identify the optimal classifier to predict student's grade level. To tackle the first problem, we investigated four sets of text complexity measures (lexical, syntactic, semantic, and character measures), some cohesion metrics and a couple of measures related to the thesis organization and to the references used on it. For the second challenge, we computed the correlation between the investigated measures and excluded the highly correlated ones and after that, we used a number of classifiers to predict the students' grade level and to compare their performances. Finally, we tested our work on a corpus of Bachelor and Master thesis from the students of the Computer Science Department of the Politehnica University of Bucharest that were written in English (as for English there is a high availability of supporting tools for natural language processing). We evaluated the quality of the presented application using Pearson's Rank Correlation to compare our obtained results with the students' grades assigned based on their thesis. | Keywords
Automatic Grading, Cohesion Metrics, Text Complexity, Natural Language Processing |
|
|
|