Lise öğrencileri için ders notlarından otomatik test oluşturma ve değerlendirme aracı

Education systems force the students to deal with numerous exams. It can be really difficult to keep motivation up while handling those exams. In today's world, information technologies have spread over to all levels of business sectors and disciplines, including education. Therefore, to make studying more target-driven and catchy, it is a good approach to use information technologies on this issue. Within this thesis, a software project, which includes the processing of text-based Turkish lecture notes of secondary education students in history and geography domains and automatic test generation, is developed. Main objective here is to provide a computer mediated self-study opportunity by simplifying an examination process with isolating students from the question preparation burden. Three question types are enabled for selection, which are true-false, fill in the blanks and multiple choice. Proposed system allows users to define constraints like number of questions and included question types before an exam generation, save generated tests or specific questions for further usage, save their exam results to keep track of their progress. Besides the educational software developed, this research draws conclusions and proposes solutions on some of the major natural language processing (NLP) tasks for Turkish like document classification, detection of sentence boundaries and headings, conversion of verbs based on their polarities after classifying them as positive and negative, constructing glossary of terms structures for history and geography domains using named entity recognition (NER) techniques.