Aiming at the sophistication of e-learning content (examination questions) for cyber security education, aiming for more appropriate automatic generation of necessary questions. In addition, we will develop teaching materials using a penetration test tool, which is a joint research with India.
WP6 ｜ Implementation report
Advancement of e-learning content for security education
Currently, a tabular database is created from the information of each component of the smart building based on the prototype of the cyber exercise course, BEMS (Building and Energy Management System) and IEEE1888, and the exercises of Moodle and MCQ are automatically performed from the database. We are developing a method to generate the database. We continued to research and develop the sophistication of e-learning content (examination questions) for cyber security education by taking the following approach.
Automatic generation of choice questions using ontology format
Through the research and development so far, the option questions for cyber security education are automatically generated by the following process.[Process of automatic generation of choice questions]
1.Model IoT device system
2. Extract keyword groups from documents such as manuals of target IoT device systems
3. By rearranging the keywords, “correct sentences” that are correct answers and “incorrect sentences” that are incorrect answers are generated.
4. Automatically generate choice questions with them as choices
In FY2019, we succeeded in improving the efficiency of automatic generation by creating a database of the relationship between the extracted keywords and the modeled IoT device system using an ontology format.
The results of this research were co-authored with Indian researchers and presented at an international conference. At the international conference, the Japanese students who mainly conducted this research and presented their research results received the Excellent Treatise Award at the international conference.
Automatic generation of fill-in-the-blank questions
We paid more attention to the fill-in-the-blank problem as e-learning content, and conducted research to automatically generate the problem. This method is performed in the following four steps.[Fill-filling problem itself 4 steps of motion generation]
1. Extract main text from manual / Create original summary text from text
2. Extract keywords (= choices for fill-in-the-blank questions) from the text
3. Create a blank sentence that serves as a template for the fill-in-the-blank question
4. Automatically generate fill-in-the-blank questions, choices, and learning materials
Here, we first conducted research on general ICT manuals and succeeded in developing a prototype.
The results of this research were co-authored with Indian researchers and presented at an international conference.
Development of teaching materials using penetration test tools
As an education for new threats, we have been developing tools that can collectively investigate vulnerabilities such as smart building using IoT devices. We continued to research and develop “IoT-PEN” , which is a framework for penetration testing.
This IoT-PEN is E2E (End-to-end), and can analyze the entire IoT device system. Compared to the conventional penetration test for each part, the IoT-PEN in this study can evaluate the resistance of the entire system to attacks, and has expandability and flexibility to large-scale and complex IoT networks. Is also shown.
The results of this research and development were co-authored with Indian researchers and presented at an international conference. In addition, the scrutinized treatise was published in the IPSJ journal of Japan. Currently, we are studying teaching materials for IoT device systems that use IoT-PEN.