Due to COVID-19, web lessons started in May 2020 at Yatsushiro Campus, National Institute of Technology (KOSEN), Kumamoto College. In order to return to face-to-face classes, it was necessary to monitor the spread of infection on campus. Therefore, we have developed a health survey system using Microsoft Forms. When students enter their body temperature and physical condition in the questionnaire, the results are posted on the campus-only website. Teachers can easily check the average body temperature and daily body temperature of students, which helps prevent infection.
In NIT Kumamoto College, Yatsushiro Campus, we had a lot of attendance book for each class before 2016. There are two types of attendance book, one created for each class and one created by each faculty member. Therefore, the total number of attendance book is 100 or more. To having a lot of attendance book makes difficulty of maintain consistency of attendance record. Solving this problem, we have developed a new computer system to manage all attendance records in 2017. In addition, COVID-19 has been confusing the world since beginning of 2020. Due to this confusion, teachers in our college use video conferencing system and/or LMS system to teach students. Unlike face-to-face lessons, teachers confirm attendance in multiple methods such as submitting report from student and/or passing quiz. Student allows submit report and/or answer quiz after class. The attendance management system also must support this method of attendance confirmation. This paper reports design, development, operation and modification of this attendance management system.
With the development of information and communication technology, various technologies and services have been born. Among them, the dialogue system is one of the technologies that have received a lot of attention. In this study, proposed non-task-oriented dialogue system is using Recurrent Neural Network (RNN) and word embedding. In dialogue systems, there are two types. One is retrieval model and the other is generative model. The dialogue system using retrieval model has few mistakes of grammar. However, it is like a parrot and takes a few seconds to generate responses. Therefore, we propose the dialogue system using generative model to improve these problems. In the proposed system, method of generating responses is seq2seq. The seq2seq is a RNN model that is trained by training data that pairs of inputs and outputs. Furthermore, the proposed system was improved to use enough vocabulary using Word2Vec. The proposed system can generate responses more natural than the retrieval model. Also, response speed was improved. Although some improvements were found, the proposed system was shown to be suitable as a dialogue system.
Botchan is a story that invites reader to a place of narrative by body expression and wordplay. Reader of Botchan becomes narratee in the place of narrative. Narratee listens to narrative of Ore. (Ore is the first person used mainly by men in Japanese). It seems that he is narrating in the style of Edokko. (Edokko has been one of the most important values since Edo period). It is full of body expression and wordplay. An analysis of his narrative shows that it is performative. He would try to make his narratee laugh by being so. Narratee can laugh at his narrative.
In the previous research, the authors clarified that blending using fly ash or ground granulated blast-furnace slag as an additive improves salinity permeation resistance compared to blending ordinary plain portland cement. In this report, with the aim of further improving salt resistance performance, special inorganic fine powders expected for high strength and high durability of concrete are used in combination with the preparation of the previous report, and salt penetration resistance is compared and evaluated. As a result, fixation of salt content was promoted by the use of the special inorganic fine powder, the apparent diffusion coefficient of chloride ions was lowered, and the improvement of salt resistance performance by the special inorganic fine powder was recognized.
We developed a new learning material that can be used in a subject of data science offered in our advanced courses. This material is for students to learn tensor data processing related to HOSVD and reflects important points to learn it. A feature of this material is that MacMahon’s cubes are used to make the students interested in the data processing. After using this material for self-study in the lecture, we confirmed degree of understanding of the matrix unfolding, the folding, and the n-mode product. As a result, it was found that about 74% of students who took the subject understood all three processes. From this, it can be considered that this material is effective for data science education of the tensor data processing.
Heavy rains in the Kumagawa River in 2020 caused extensive damage to its watershed. Meanwhile, the “Hachinoji” weir constructed downstream of the Kumagawa River has been maintained, including a park in the vicinity, but the floods this time destroyed the park as well. This paper shows the damage to this weir and its surrounding area, and it was found that the committee needs to discuss the direction of the future utilization of the high-water channel.
The torrential rain that occurred in July 2nd year of Reiwa caused enormous damage mainly in Kumamoto prefecture. Many river floods occurred in the Kuma River. We conducted a field survey after this disaster. We confirmed overflow 5 places in Hitoyoshi City. Two of these were dike break. And, we confirmed water leakage 1 place in Yatsushiro City. The two parts of overflow have similar characteristics. The damaged parts are all discontinuous in embankment longitudinal direction.
This paper is a report of the disaster investigation of July Heavy rain in Reiwa 2nd. This heavy rains have drained many bridges over the Kuma River. There is no fact that so many bridges over class a river have been flowed out in past disasters. It is important to investigate the damaged bridges and analyze the outflow mechanism for future disaster countermeasures. This paper shows the bridge disaster survey conducted from the time of the disaster to August. And it shows some research points for the future.
We have applied various 3D puzzles to the education of multi dimensional data processing and developed online learning material for data science programming based on them. In addition, it also uses a Learning Management System (LMS). In recent years, active learning (AL) style class have been attracting attention. The aims of this style are not only gaining knowledge but also understanding how to apply knowledge and realizing importance of active attitude for study. In our school, a part of math classes is introduced this style which students spent most of time for problem exercise. In addition, we have to offer classes using internet under the influence of COVID-19 from May 2020. In this paper, we explain how to conduct the face-to-face math class introduced AL style and how to apply LMS to this class, including online. Furthermore, we describe the merits and demerits founded by this practice.
In this paper, we applied structural equation modeling (SEM) and latent growth model (LGM) to evaluate the effects of vital signs and environmental data, which were obtained during 90 minutes after falling asleep, on sleep quality. The data used for the analysis is obtained by a care support device used in an elderly care facility. Applying SEM analysis to the data 90 minutes after falling asleep revealed factors that deepen sleep. Then applying LGM analysis, we found that each sleep level factors have a different sleep-deepening effect. Statistical analysis environment R and the lavaan package were used for both SEM and LGM analysis in this paper. The results show that SEM and LGM can be used to build a rational model of the effects of vital signs and environmental conditions on deepening sleep.
We developed low-cost wireless device for face-to-face teaching support to get responses of each students in classroom. It contains four push-switches, four LEDs, wireless transmitter/receiver module ESP-WROOM-32 and two AA batteries. We call it “YONTAKUN’’ (Yielding Online Numbered Touch Actions for Knowledge Understanding Notifier). Fifty of those devices were duplicated and used on a trial base. It was designed to tell us not only which of the four switches were pressed, but also that multiple switches were pressed at the same time and the order in which they were pressed slowly. It can be used to confirm students' pre- or post-learning knowledge while displaying the results of the questionnaire at the same time.
KH Coder, which is a powerful free software for text mining, has a feature of a user-defined dictionary to adopt particular needs of users. In the report, a user-defined dictionary is created for content analysis of questions of the national examination for Technical Radio Operator for On-the-Ground Services in Japan: Technical terms necessary for learning radio engineering are extracted from all the questions in the national examinations that were ever carried out 36 times and then the technical terms are registered into the user-defined dictionary. Two examples of the content analysis with the user-defined dictionary are included.