Eye tracking provides a marker of attention. In the educational context, such behavior can be harnessed to understand learning behaviors. However, a technology framework that captures and utilizes such multimodal indicators in educational activities is lacking. This paper presents LA-ReflecT, a platform integrating multimodal data for micro-learning activities. Teachers can author learning tasks and enable tracking eye fixation behaviors. A web camera-based eye-tracking function captures the gaze data while attempting the learning task. Learners can control the settings to stop or pause recording. We present data-driven services such as visualizing gaze attention heatmap and genetic algorithm-based group formation. A classroom study with 41 students illustrates using the proposed framework in an authentic context. Data collected is analyzed to answer an initial research question regarding the correlation between the heterogeneity of the click and gaze patterns in a learning task. The work is open for a demo.

Post-exam reflection is critical in helping students consolidate knowledge acquired during a course, enabling them to apply this understanding in future professional contexts. This study investigates the effectiveness of Mirai, a large language model-based (LLM) chatbot, in supporting students' post-exam reflection needs in an Automation Engineering course. Through a controlled experiment, we explored how context-tuned and non-context-tuned versions of Mirai impacted students' reflection habits, help-seeking behaviors, and perceptions of the tool. Students interact with the chatbot to clarify exam questions and receive personalized explanations. A a surveys based on the extended technology acceptance model (exTAM) was conducted and the resulting data was analyzed. We assessed the efficacy of Mirai in facilitating a deeper understanding of exam-related material, improving students' knowledge, engagement and performance. The findings from this study provide insights into the immediate educational benefits of LLM-based tools, their acceptance among students, and their role in enhancing learning outcomes in engineering education.

The rapid development of Generative AI (GenAl) provides more opportunities and methods to deliver meaningful, engaging and gamified learning experiences to language learners. While there are various language learning applications, current methods often suffer from low completion rates and a painful learning process. In this paper, we propose a new gamified learning experience for English Language learners based on an image-text-image GenAl game: AVERY (Augmenting Vision to Enhance YouR English writing skills). The game is designed to enhance learner engagement by adopting image generation in English writing. A learner begins by providing the system with an image. The learner can ask the AI for hints to describe the image and pass a well-curated sentence to the system. The system generates an image based on the learner's answer. In the final round, the system provides feedback on how well the learner provided useful and correct clues and areas for further improvement. 12 respondents were asked to play the game and fill a questionnaire. The results showed a positive affect towards the AVERY system and its use in enhancing learner engagement.

Learning EFL through translation tasks is an effective language learning technique, but requires consistent practice and scaffolding. This study evaluates TAMMY, a prototype EFL chatbot designed for Japanese learners to practise English translation tasks. Using the extended Technology Acceptance Model, the study examines Tammy's usability, usefulness, and enjoyment. Response appropriateness and task success are also explored. Findings from a pilot study with Japanese university students indicate high usability and positive attitudes towards the chatbot. Tammy effectively provided accurate feedback in most tasks successfully guiding learners to an accurate translation, though improvements are needed in feedback clarity and conversational adaptability. Despite limitations, Tammy shows promise as a support tool for language learning, offering an engaging and non-judgmental platform for practising translation and enhancing English proficiency.

Learning English as a foreign language is a core part of K-12 education for many countries in which English is not the main spoken language, and especially in Asia. One of the fundamental tasks that students encounter is to learn vocabulary that is a part of the assigned curriculum. These are often sourced from reference materials or assigned vocabulary lists and may not consider the learner’s current proficiency or the semantic context of words that were recently learnt. By suggesting vocabulary that have similar proficiency or semantic contexts to what a student has recently studied could improve and support vocabulary learning. In this paper, we propose a method for recommending words that have similar difficulty and semantic context with previous words learnt based on the analysis of prescribed textbooks for Japanese junior high school students. This research could be used to guide a student learning English by helping them select a sequence of vocabulary that is appropriate.

As Learners move from one learning environment to another, there is a key necessity of taking with them a proof of previous learning achievements or experiences. In most cases, this is either expressed in terms of receipt of scores or a certificate of completion. While this may be sufficient for enrollment and other administrative decisions, it poses some limitations to the depth of learning analytics and consequently a slow onboarding process. Also, with different institutions having their learning data isolated from each other, it becomes more difficult to easily access a learner's learning history for all learning activities on other systems. In this paper, we propose a blockchain based approach for connecting learning data across different Learning Management Systems (LMS), Learning Record Stores (LRS), institutions and organizations. Leveraging on unique properties of blockchain technology, we also propose solutions to ensuring learning data consistency, availability, immutability, security, privacy and access control.