![]() | Only 14 pages are availabe for public view |
Abstract Text similarity is a basic block in many NLP applications. While there exists a wide spectrum of applications that utilizes the text similarity techniques, what gathered our attention is the utilization of text similarity in the academic domain. Text similarity has been the backbone in many applications, i.e. (Automatic scoring, plagiarism detection, machine translation, search engine results ranking etc.). In this thesis two applications are tackled (short answer grading and Arabic plagiarism detection). Alongside providing an intensive analysis for both tasks, two systems are proposed. The proposed systems achieved very high accuracy compared to state of the art systems for both tasks |