Learning Text-image Joint Embedding for Efficient Cross-modal Retrieval with Deep Feature Engineering

This article introduces a two-phase deep feature engineering framework for efficient learning of semantics enhanced joint embedding, which clearly separates the deep feature engineering in data preprocessing from training the text-image joint embedding model. We use the Recipe1M dataset for the technical description and empirical validation. In preprocessing, we perform deep feature engineering by combining deep feature engineering with semantic context features derived from raw text-image input data. We leverage LSTM to identify key terms, deep NLP models from the BERT family, TextRank, or TF-IDF to produce ranking scores for key terms before generating the vector representation for each key term by using Word2vec. We leverage Wide ResNet50 and Word2vec to extract and encode the image category semantics of food images to help semantic alignment of the learned recipe and image embeddings in the joint latent space. In joint embedding learning, we perform deep feature engineering by optimizing the batch-hard triplet loss function with soft-margin and double negative sampling, taking into account also the category-based alignment loss and discriminator-based alignment loss. Extensive experiments demonstrate that our SEJE approach with deep feature engineering significantly outperforms the state-of-the-art approaches.

Non-millennial teachers’ strategies in coping with the online teaching during the COVID-19 pandemic

This article discusses the strategies used by five non-millennial teachers (aged 54-59 years old) of a junior high school in coping with the online teaching during the COVID-19 pandemic, in Ngawi, East Java, Indonesia. The teachers were interviewed, and the data were transcribed and analyzed by creating a data repository, expanding the codes, describing the coded data, and drawing conclusions. The results revealed that the teachers had several strategies used, which were to increase students’ interest in learning, provide students with knowledge and attention, create efficient learning resources, and use SIMPEL (Sistem Informasi Manajemen Pembelajaran or Learning Management Information System), which is specifically available only in Ngawi. SIMPEL was specially developed by the Ngawi district education office, to ensure that the learning processes in Ngawi Regency continue to run optimally during the COVID-19 outbreak. SIMPEL substituted the use of online YouTube videos and materials because the materials were already provided by the system, decreasing the need for the teachers to depend on other resources. Despite these teachers also using other online platforms, hence issues such as the slow internet connection, running out of quotas and blackouts, hindered their efforts to use these platforms at times. Hence, WAG was the most used media to conduct their online learning due to its simplicity and availability. These teachers continued to strive to learn digital technologies ever since they changed from their previous face-to-face teaching strategies.

Construction of a Framework for Selecting an Effective Learning Procedure in the School-Level Sector of Online Teaching Informatics

Currently, there is no way soon to stop the coronavirus epidemic that has spread over the globe. People are alarmed by its quick and widespread expansion. COVID-19's transmission chain was then broken by everyone. There was a gradual decrease in social and physical closeness. Distancing yourself from others is a way to prevent the transmission of disease. The purpose of this research is to investigate how online learning can be implemented in Tamil Nadu, India, during the COVID-19 epidemic. This research works focuses to find efficient learning procedure in eLearning protocols. The findings indicated that Google Classroom, WhatsApp, and Zoom Clouds Meeting were consecutively the most commonly utilized programs to help in remote learning. Despite this, most instructors continue to use the learning paradigm while teaching in virtual environments. Online learning and remote education are the most common methods of learning. The instructor claims that the learning model used is beneficial to their work in creating a virtual classroom since it adheres to the model's structured grammar. The experimental test has been conducted with 125 students who anonymously filled out a questionnaire and voted for more visual based eLearning. The findings show that students in distance education believed that there were more tasks than in face-to-face education. At the same time, students indicated that they spent more time studying at home than in school.

Use of simulation-based learning in Japanese undergraduate nursing education: National survey results

Introduction: Simulation-based learning (SBL) is a practical and efficient learning method that involves the replacement of a portion of clinical education with quality simulation experiences. It has been utilised in various countries, such as the United States, Canada, and South Korea. However, based on current regulations in Japan, clinical education cannot be replaced with simulation experience. For future curriculum integration, it is necessary to clarify the current use of SBL and tackle systematic educational strategies of SBL. Therefore, this national survey aimed to clarify the prevalence and practices of SBL in undergraduate nursing education programs in Japan. Methods: This article presents the results of our national survey in Japan. It presents the questionnaire based on the International Nursing Association for Clinical Simulation and Learning Standards of Best Practice and demonstrates the use of simulation-based learning in Japanese undergraduate nursing programs. Results: Overall, the schools using simulation-based education (SBE) comprised 346 schools (82.4%) of the sample. Those equipped with high-fidelity simulators were 146 schools (27.6%); the rest owned medium-fidelity simulators. Almost all undergraduate nursing education systems were equipped with simulators, however, the frequency of use was low. SBL was incorporated into the curriculum at many undergraduate nursing education institutions, and awareness of the INACSL Standard of Best Practice: SimulationSM was extremely low. Conclusion: This study shows that SBL is not properly utilised in undergraduate nursing programs, even though many schools are equipped with simulators. Thus, further study on barriers to simulator use is needed.

Efficient learning in children with rapid GABA boosting after training

It is unclear why and how children learn more efficiently than adults, although inhibitory systems, which play an important role in stabilizing learning, are immature in children. Here, we found that despite a lower baseline concentration of gamma-aminobutyric acid (GABA) in early visual cortical areas in children (8 to 11 years old) than adults (18 to 35 years old), children exhibited a rapid boost of GABA immediately after visual training, whereas the concentration of GABA in adults remained unchanged after training. Moreover, behavioral experiments showed that children stabilized visual learning much faster than adults, showing rapid development of resilience to retrograde interference. These results together suggest that inhibitory systems in children's brains are more dynamic and adapt more quickly to stabilize learning than in adults.

IMPLEMENTASI PENDEKATAN SAINTIFIK PADA PEMBELAJARAN ILMU PENGETAHUAN ALAM DI MIS IKHWANUL MUKMININ

The implementation of a scientific approach is something that needs to be studied more deeply in order to create an active and efficient learning process. This study aims to determine the implementation of learning plans and the application of the scientific approach by science teachers. The method used in this research is a qualitative research method which is also known as naturalistic research. The result of this study is that the science teacher understands the theoretical scientific approach, and is equipped with a plan in the RPP document which is prepared before the lesson takes place. In the process of applying the scientific approach, it takes a longer time to carry out observations, where the media displayed is limited to the use of simple paper, due to the limited supporting learning media such as LCD projectors, and so on.

Digital Communication in the Inclusive Classroom

Digital communication is being extensively used, and during COVID-19 pandemic, it has transformed the way teaching is delivered and how learning happens which became even more problematic for children with learning difficulties. The digitalization of education during the lockdown period has forced teachers, children, and parents to develop and enhance their digital skills to maintain and keep ensuring efficient learning. Digital communication can be provided in the educational system by the simple use of email or WhatsApp groups up to the integration of complex digitalized learning programs and software adapted to the specific educational needs of each student. Being digital natives, students nowadays seem more engaged if learning is mediated by the use of digital communication tools. They are opened and interested in participating in educational activities that are technology-based.

Entropy-Based Variational Scheme with Component Splitting for the Efficient Learning of Gamma Mixtures

Finite Gamma mixture models have proved to be flexible and can take prior information into account to improve generalization capability, which make them interesting for several machine learning and data mining applications. In this study, an efficient Gamma mixture model-based approach for proportional vector clustering is proposed. In particular, a sophisticated entropy-based variational algorithm is developed to learn the model and optimize its complexity simultaneously. Moreover, a component-splitting principle is investigated, here, to handle the problem of model selection and to prevent over-fitting, which is an added advantage, as it is done within the variational framework. The performance and merits of the proposed framework are evaluated on multiple, real-challenging applications including dynamic textures clustering, objects categorization and human gesture recognition.