Web e-Learning: Automated Essay Assessment Based on Natural Language Processing Using Vector Space Model

Automatic essay assessment aims to determine the value of essay and provides feedback based on a computer system. Using a computer-based system can make essay assessments more objective so that the students are more satisfied with the results they get. The assessment of the learning outcomes is very important for web e-learning systems. The application of the right algorithm can provide the right assessment results. This article aims to develop a web e-Learning system with the automatic essay answer scoring application based on natural language processing (NLP) with the text processing. It consists of tokenizing, stopward removal and stemming by using a vector space model algorithm to calculate the word weights. The feasibility validation test phase of application system was carried out by comparing between manual assessment and automatic assessment. The results of manual calculation shows the similaritty with a percentage of 82%. while the level of similarity of the automatic calculation system is 81.6%. The calculation value of similarity to all sample questiosn tested has a significant proximity.