Automated Student Assignment Evaluation Based on Information Retrieval and Statistical Techniques

Efficient automated evaluation is urgently needed in programming courses within the field of online education. This study introduces a model that focuses on the TF-IDF algorithm, a reliable technique for information retrieval, to simplify the assessment procedure in programming courses with ten distinct categories. TF-IDF is a numerical statistical technique that measures the relevance of tokens. When combined with cosine similarity, it yields impressive outcomes in document scoring and ranking. The suggested paradigm greatly decreases the amount of time and effort teachers need to spend on manual assessment. The experimental results confirm the effectiveness of the model, demonstrating its strong ability to automatically assess programming tasks.