Search Results (30)

Machine learning in natural language processing (NLP) analyzes datasets to make future predictions, but developing accurate models requires large, high-quality, and balanced datasets. However, collecting such datasets, especially for low-resource languages, is time-consuming and costly. As a solution, data augmentation can be used to increase the dataset size by generating synthetic samples from existing data. This study examines the effect of translation-based data augmentation on sentiment analysis using small datasets in three diverse languages: French, German, and Japanese. We use two neural machine translation (NMT) services—Google Translate and DeepL—to generate augmented datasets through intermediate language translation. Sentiment analysis models based on Support Vector Machine (SVM) are trained on both original and augmented datasets and evaluated using accuracy, precision, recall, and F1 score. Our results demonstrate that translation augmentation significantly enhances model performance in both French and Japanese. For example, using Google Translate, model accuracy improved from 62.50% to 83.55% in Japanese (+21.05%) and from 87.66% to 90.26% in French (+2.6%). In contrast, the German dataset showed a minor improvement or decline, depending on the translator used. Google-based augmentation generally outperformed DeepL, which yielded smaller or negative gains. To evaluate cross-lingual generalization, models trained on one language were tested on datasets in the other two. Notably, a model trained on augmented German data improved its accuracy on French test data from 81.17% to 85.71% and on Japanese test data from 71.71% to 79.61%. Similarly, a model trained on augmented Japanese data improved accuracy on German test data by up to 3.4%. These findings highlight that translation-based augmentation can enhance sentiment classification and cross-language adaptability, particularly in low-resource and multilingual NLP settings. Full article

While Japanese universities promote English proficiency and intercultural competence, admissions still rely heavily on the standardised TOEIC test, often prioritising linguistic accuracy over cultural understanding. This emphasis can marginalise the experiential and interpersonal aspects of language learning essential for developing intercultural communicative competence (ICC). Although ICC is most effectively developed through a combination of formal instruction and informal learning, particularly involving authentic intercultural interactions, there remains a paucity of research examining how learners develop ICC beyond the classroom. Addressing this gap, the present study investigates how Japanese university students’ language repertoires, intercultural experiences, and knowledge shape their studies, and how their ICC develops throughout undergraduate education. One English major student was selected for in-depth analysis following his employment in a culturally diverse workplace that fostered meaningful intercultural interactions. His development was examined using four research tools: a survey, a journal, an interview, and TOEIC scores. The survey and test scores revealed growth in communicative assurance and language proficiency, while journal reflections and interviews captured critical incidents—including a shift from indirect to direct speech—that marked cultural adaptation. These findings suggest experiential learning fosters real-time communication, builds learner confidence, and supports transformational ICC development. Full article

Recent advances in large multimodal models (LMMs) have opened new opportunities for multitask recognition from remote sensing images. However, existing approaches still face challenges in effectively recognizing the complex geospatial characteristics of regions such as Japan, where its location along the seismic belt leads to highly diverse urban environments and cityscapes that differ from those in other regions. To overcome these challenges, we propose the GeoJapan Fusion Framework (GFF), a multimodal architecture that integrates a large language model (LLM) and a vision–language model (VLM) and strengthens multimodal alignment ability through an in-context learning mechanism to support multitask recognition for Japanese remote sensing images. The GFF also incorporates a cross-modal feature fusion mechanism with low-rank adaptation (LoRA) to enhance representation alignment and enable efficient model adaptation. To facilitate the construction of the GFF, we construct the GeoJapan dataset, which comprises a substantial collection of high-quality Japanese remote sensing images, designed to facilitate multitask recognition using LMMs. We conducted extensive experiments and compared our method with state-of-the-art LMMs. The experimental results demonstrate that GFF outperforms previous approaches across multiple tasks, demonstrating its promising ability for multimodal multitask remote sensing recognition. Full article

To improve the accuracy of Japanese finger-spelled character recognition using an RGB camera, we focused on feature design and refinement of the recognition method. By leveraging angular features extracted via MediaPipe, we proposed a method that effectively captures subtle motion differences while minimizing the influence of background and surrounding individuals. We constructed a large-scale dataset that includes not only the basic 50 Japanese syllables but also those with diacritical marks, such as voiced sounds (e.g., “ga”, “za”, “da”) and semi-voiced sounds (e.g., “pa”, “pi”, “pu”), to enhance the model’s ability to recognize a wide variety of characters. In addition, the application of a change-point detection algorithm enabled accurate segmentation of sign language motion boundaries, improving word-level recognition performance. These efforts laid the foundation for a highly practical recognition system. However, several challenges remain, including the limited size and diversity of the dataset and the need for further improvements in segmentation accuracy. Future work will focus on enhancing the model’s generalizability by collecting more diverse data from a broader range of participants and incorporating segmentation methods that consider contextual information. Ultimately, the outcomes of this research should contribute to the development of educational support tools and sign language interpretation systems aimed at real-world applications. Full article

Background/Objectives: Diagnostic reasoning is essential in clinical practice and medical education, yet it often becomes an automated process, making its cognitive mechanisms less visible. Despite the widespread use of electronic medical records, few studies have quantitatively evaluated how clinicians’ reasoning is documented in real-world electronic medical records. This study aimed to investigate whether initial electronic medical records contain valuable information for diagnostic reasoning and assess the feasibility of using text analysis and logistic regression to make this reasoning process visible. Methods: We conducted a retrospective analysis of initial electronic medical records at Kochi University Hospital between 2008 and 2022. Two patient cohorts presenting with dizziness and headaches were analysed. Text analysis was performed using GiNZA, a Japanese natural language processing library, and logistic regression analyses were conducted to identify associations with final diagnoses. Results: We identified 1277 dizziness cases, of which 248 were analysed, revealing 48 significant diagnostic terms. Moreover, we identified 1904 headache cases, of which 616 were analysed, revealing 46 significant diagnostic terms. The logistic regression analysis demonstrated that the presence of specific terms, as well as whether they were expressed affirmatively or negatively, was significantly associated with diagnostic outcomes. Conclusions: Initial EMRs contain quantifiable linguistic cues relevant to diagnostic reasoning. Even simple analytical methods can reveal reasoning patterns, offering valuable insights for medical education and supporting the development of explainable diagnostic support systems. Full article

A recent study found evidence of symbolic gesture use in Japanese tits (Parus minor). The study reveals how these birds use wing-fluttering movements to transmit an “after you” directive to their partners, implying a degree of cognitive skill previously thought to be unique to humans and great apes. If confirmed, this research contradicts long-held notions about animal communication by proving that Japanese tits not only participate in intricate vocal communications, which can comprise phrases with specific grammatical rules, but also use body language as a form of engagement. Here, we evaluate this claim by inspecting the data and utilizing bootstrapping to expand the sample size. We find a large variation in how frequently the “after you” gesture is employed in different bootstrap samples, suggesting that while the gesture is a consistent behavior, its frequency can vary significantly. Moreover, the timing of the male’s response to the female’s gesture can fluctuate, though the response itself appears to be a stable phenomenon. Beyond evaluating the bootstrapping technique, we employ causal inference to examine whether wing fluttering by female Japanese tits leads to a quicker response by males, entering the nest more swiftly. The analysis supports this hypothesis, suggesting that wing fluttering functions as an effective communicative gesture in this species, influencing male nesting behavior. However, because the frequency of the “after you” gesture varied greatly, especially in light of potential cognitive and other biases influencing the study, this bold claim should be taken with caution. Full article

In Japan, where the shortage of care workers poses a challenge to the sustainability of the long-term care system, foreign care workers play a crucial role. This study aimed to identify the needs and challenges of Vietnamese Technical Intern Training candidates, the largest group of foreign care workers, to facilitate program reforms and a more efficient recruitment process. A semi-structured interview survey was conducted with 27 candidates in Vietnam. Interview items included reasons for choosing the training program in Japan, the desired length of stay, expectations, and career advancement after returning home. Descriptive statistics and K-means clustering were used to analyze the data. Survey results showed that all participants independently decided to pursue care worker training in Japan; 44% had considered other countries; most wanted to stay in Japan for as long as possible; and 37% wanted to live in Japan permanently. The K-means method revealed three clusters: a Japanophile cluster (preferred Japan for its landscape, culture, and national character); a word-of-mouth cluster (influenced by personal referrals); and an intellectual cluster (influenced by Japan’s economic development and care levels). Our findings indicate that support for obtaining qualifications, Japanese language skills, and caregiving skills are important to secure the stability of foreign care workers. It is necessary to tailor recruitment, training, and support for each cluster. Full article

Multi-word expressions make up a large proportion of the English language and particularly spoken language. Using multi-word expressions can assist with the impression of fluency, making them useful for language learners to know and use. However, proven methods for teaching this language phenomenon are required, so that learners can easily use multi-word expressions in their conversations. The purpose of the study was to examine the efficacy of a fluency workshop focused on multi-word expression use in conversation and to determine the most appropriate implementation for the Japanese context. An action research structure was used over three iterations of the fluency workshop, learner feedback and teacher observations were used to make improvements. Learner feedback regarding the usefulness of each activity for learning English was compared between the original cohort and subsequent cohorts. The results showed significant differences in levels of perceived usefulness for activities where improvements were made, but also for some activities where no specific improvement was made, suggesting that teaching improves through practice. Pedagogical implications include maximising the time on task via clear instructions, providing visual time constraints, and offering scaffolding to support the use of multi-word expressions when recall seems beyond a learner. Full article

Japan is experiencing significant demographic shifts due to an aging and declining population. In 1993, the Japanese Government introduced the Technical Intern Training Program (TITP) to accept foreign national workers. While the number of technical intern trainees under this program has constantly increased, many of them face challenges in their daily lives, such as stress, health insecurities and limited access to healthcare. Therefore, we conducted a cross-sectional study to assess the mental well-being of technical intern trainees, focusing on psychological distress and its related factors. This study included 304 technical intern trainees from 12 prefectures in Japan, and was conducted from August 2021 to October 2021. We used self-administered questionnaires in the participants’ native languages to assess their sociodemographic conditions, health-related conditions such as health insecurities and healthcare-seeking behaviors, and psychological distress. The K6 Kessler Psychological Distress Scale was applied to evaluate the levels of psychological distress. Among the participants, 26.3% had moderate psychological distress and 2.3% had severe psychological distress. In addition, about 15% of the participants reported health insecurities and did not see a doctor despite wanting to. The multivariate model of logistic regression revealed significant associations between psychological distress and female gender (AOR 2.62, 95% CI 1.12–6.12), nationality other than Vietnamese (AOR 4.85, 95% CI 2.60–9.07), tough financial conditions (AOR 2.23, 95% CI 1.18–4.19), experiencing health insecurity (AOR 2.21, 95% CI 1.04–4.66) and the health behavior of refraining from seeking medical care (AOR 3.06, 95% CI 1.49–6.30). The top reasons for refraining from seeking medical care were the participants’ limited knowledge about the healthcare system in Japan and their worries about medical bills. These findings highlight the necessity to extend mental health support services, including counseling services, and share information about Japan’s healthcare system to supply medical services to the targeted technical intern trainee population. Full article

Growing attention has been devoted to the contribution of morphological knowledge to reading comprehension. Because of the complex nature of morphological knowledge, more fine-grained approaches are sought on this topic by exploring multiple aspects of morphological knowledge and multiple pathways through which each aspect contributes to reading comprehension. This study measured three aspects of affix knowledge (form, meaning, and use) and vocabulary breadth and examined how each aspect contributes to EFL (English as a foreign language) reading comprehension by modeling direct and indirect effects with vocabulary as a mediator. The participants were 211 Japanese university students. All variables were measured using standardized tests. Direct effects of meaning, use, and vocabulary and indirect effects of meaning and use via vocabulary were significant. However, form displayed no significant effect. The lack of significant effects for form may be due to the design of this study, which did not include word reading (a variable that may mediate form’s effect). In sum, although the form aspect did not show any effect, semantic and syntactic aspects demonstrated direct and indirect contributions. Overall, this study endorsed the criticality of a more fine-grained approach, shedding light on what and how morphological knowledge supports L2 reading comprehension. Full article

Japanese Sign Language (JSL) is vital for communication in Japan’s deaf and hard-of-hearing community. But probably because of the large number of patterns, 46 types, there is a mixture of static and dynamic, and the dynamic ones have been excluded in most studies. Few researchers have been working to develop a dynamic JSL alphabet, and their performance accuracy is unsatisfactory. We proposed a dynamic JSL recognition system using effective feature extraction and feature selection approaches to overcome the challenges. In the procedure, we follow the hand pose estimation, effective feature extraction, and machine learning techniques. We collected a video dataset capturing JSL gestures through standard RGB cameras and employed MediaPipe for hand pose estimation. Four types of features were proposed. The significance of these features is that the same feature generation method can be used regardless of the number of frames or whether the features are dynamic or static. We employed a Random forest (RF) based feature selection approach to select the potential feature. Finally, we fed the reduced features into the kernels-based Support Vector Machine (SVM) algorithm classification. Evaluations conducted on our proprietary newly created dynamic Japanese sign language alphabet dataset and LSA64 dynamic dataset yielded recognition accuracies of 97.20% and 98.40%, respectively. This innovative approach not only addresses the complexities of JSL but also holds the potential to bridge communication gaps, offering effective communication for the deaf and hard-of-hearing, and has broader implications for sign language recognition systems globally. Full article

Simultaneous bilingual and bicultural children who are schooled in the dominant societal language can acquire literacy in their home language through home literacy practices and weekend school. Twenty-eight Japanese-English bilingual–bicultural children (ages 9 to 14) attending English weekend schools in Japan were assessed using the standardized Test of Written Language (TOWL, 4th ed.). Their overall age-adjusted writing scores showed that most were on a par with same-age US peers. The parents’ responses to parental questionnaires revealed that the children’s regular attendance and enjoyment of weekend school, coupled with parental home support and access to plenty of appropriate English reading materials, helped to establish their English literacy. Nevertheless, the children attained different writing scores. Although many performed within the Average range, some scored either in the Above Average or higher range or the Below Average or lower range. Further analysis of these three groups of children revealed a strong relationship between reading practices and writing scores. Better writers tend to read frequently in English; they enjoy reading independently and voluntarily. These findings underscore the importance of recreational reading as an accessible and affordable means for developing home language literacy in bilingual and bicultural children in Japan. Full article

This study examines experiences and perceptions concerning the provision of Japanese language support for ethnic minority school children between the ages of 6 and 12 enrolled in public elementary schools in a city located in the Kansai region of Japan (hereinafter referred to as “City M”). This paper will focus in particular on interpreting the experiences and perspectives of language support teachers, volunteer interpreters, mother-tongue language supporters (hereafter referred to as MTLS) as well as three principals of three public elementary schools located in particularly different areas of the City M. Each of the school’s history, backgrounds and current situations are varied and unique. One-on-one interviews of 40 to 60 min in length were conducted with a total of 9 participants consisting of public elementary school principals and Japanese language support teachers. These results of the data collection provided a deeper understanding and explanation of the reasons behind current trends and challenges regarding the accessibility, implementation, provision, and practice of language learning support for ethnic minority school children. The findings from this research will increase awareness of current issues faced by practitioners supporting ethnic minority children in their learning. The research findings also provide insight into what obstacles need to be overcome in order to provide adequate, sufficient, and sustainable educational support for children from different cultural and linguistic backgrounds within the mainstream Japanese education system. Full article

Research in study abroad (SA) has developed over the past several decades, with an increasing focus on individual learners and their experiences. Despite this trend, it remains unclear what affordances beyond the classroom are available to students and whether and how learners enact their agency to make use of the available opportunities. The present study adopts an ecological framework to examine how four Japanese learners enacted their agency to capitalize on affordances for language learning beyond the classroom during study abroad. Agency is understood as a multidimensional construct, with both internal and external dimensions. Learners utilized a smartphone application to regularly report their language use and associated reflections. Further details about their experiences were elicited by post-study abroad questionnaires and interviews. The findings reveal the various kinds of language episodes reported by the students and how these learners exercised their agency in relation to their learning experiences. A discussion of what factors likely contributed to these learners’ enacting their agency is included. The study concludes with implications for supporting students’ out-of-class learning during study abroad. Full article

There are various mobile applications to support foreign-language learning. While providing interactive designs and playful games to keep learners interested, these applications do not focus on motivating learners to continue learning after a long time. Our goal for this study was to develop an application that guides learners to achieve small goals by creating small lessons that are related to their real-life situations, with a main focus on vocabulary acquisition. Therefore, we present MiniHongo, a smartphone application that recognizes learners’ current locations and activities to compose lessons that comprise words that are strongly related to the learners’ real-time situations and can be studied in a short time period, thereby improving user motivation. MiniHongo uses a cloud service for its database and public application programming interfaces for location tracking. A between-subject experiment was conducted to evaluate MiniHongo, which involved comparing it to two other versions of itself. One composed lessons without location recognition, and the other composed lessons without location and activity recognition. The experimental results indicate that users have a strong interest in learning Japanese with MiniHongo, and some difference was found in how well users could memorize what they learned via the application. It is also suggested that the application requires improvements. Full article