Papers by Sinem Emine Mete
Educational Technology Research and Development, 2018
The goal of this research is to investigate the effect of emotion-aware interventions on students... more The goal of this research is to investigate the effect of emotion-aware interventions on students' behavioral and emotional states. To this end, we collected data from 12 students in the 9th grade in a high school in Turkey. The data collection took place in two sessions of an English Course. While the students were reading articles and solving relevant questions, our data collection application running in the background recorded the videos of the individual students through a camera and captured students' screens in a non-intrusive manner. In total, we had 12.5 h of student data. We employed the human expert labeling process (HELP) (Aslan et al. in Workshop proceedings at international conference on intelligent tutoring systems (ITS), pp 156-165, 2016) to have the data labeled (150 h of data labeling in total). The data collection application was designed in a way that it also collected emotional self-labels (i.e., emotional states as self-reported by students at any time of learning). We leveraged emotional self-label information to suggest various real-time interventions for the students. The results obtained using the final expert labels showed that the percentage of the students' Satisfied state was significantly higher after interventions. The results also demonstrated that although the interventions were triggered by the emotional states as self-labeled by the students and tailored to improve such states, there was a major positive impact of these interventions on students' behavioral states. This preliminary study showed that even with a limited set of emotion-aware interventions based on self-labels, students' states could be impacted positively. Conducting large-scale pilots leveraging more advanced interventions is a future direction for our research.
Bookmarks Related papers MentionsView impact
CHI, 2019
We developed a real-time, multimodal Student Engagement Analytics Technology so that teachers can... more We developed a real-time, multimodal Student Engagement Analytics Technology so that teachers can provide just-in-time personalized support to students who risk disengagement. To investigate the impact of the technology, we ran an exploratory semester-long study with a teacher in two classrooms. We used a multi-method approach consisting of a quasi-experimental design to evaluate the impact of the technology and a case study design to understand the environmental and social factors surrounding the classroom setting. The results show that the technology had a significant impact on the teacher's classroom practices (i.e., increased scaffolding to the students) and student engagement (i.e., less boredom). These results suggest that the technology has the potential to support teachers' role of being a coach in technology-mediated learning environments. CCS CONCEPTS • Applied computing-> Education-> Learning management systems • Human-centered computing-> Human computer interaction (HCI)-> Empirical studies in HCI
Bookmarks Related papers MentionsView impact
Educational Technology Research and Development, 2018
Bookmarks Related papers MentionsView impact
Although there are some implementations towards understanding students' emotional states through ... more Although there are some implementations towards understanding students' emotional states through automated systems with machine learning models, one of the key challenges still remain unaddressed: Generic detectors of emotions lack enough accuracy to autonomously and meaningfully trigger any interventions to infuse positive change in students. Collecting self-labels from students as they assess their internal states can be a way to collect labeled subject specific data necessary to obtain personalized emotional engagement models. In this paper, we outline preliminary analysis on emotional self-labels collected from students while using a 1:1 math learning platform.
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Affective states play a crucial role in learning. Existing Intelligent Tutoring Systems (ITSs) fa... more Affective states play a crucial role in learning. Existing Intelligent Tutoring Systems (ITSs) fail to track affective states of learners accurately. Without an accurate detection of such states, ITSs are limited in providing truly personalized learning experience. In our longitudinal research, we have been working towards developing an empathic autonomous 'tutor' closely monitoring students in real-time using multiple sources of data to understand their affective states corresponding to emotional engagement. We focus on detecting learning related states (i.e., 'Satisfied', 'Bored', and 'Confused'). We have collected 210 hours of data through authentic classroom pilots of 17 sessions. We collected information from two modalities: (1) appearance, which is collected from the camera, and (2) context-performance, that is derived from the content platform. The learning content of the content platform consists of two section types: (1) instructional where students watch instructional videos and (2) assessment where students solve exercise questions. Since there are individual differences in expressing affective states, the detection of emotional engagement needs to be customized for each individual. In this paper, we propose a hierarchical semi-supervised model adaptation method to achieve highly accurate emotional engagement detectors. In the initial calibration phase, a personalized context-performance classifier is obtained. In the online usage phase, the appearance classifier is automatically personalized using the labels generated by the context-performance model. The experimental results show that personalization enables performance improvement of our generic emotional engagement detectors. The proposed semi-supervised hierarchical personalization method result in 89.23% and 75.20% F1 measures for the instructional and assessment sections, respectively.
Bookmarks Related papers MentionsView impact
Educational Technology, 2016
In this article the authors start with a description of the learner-centered paradigm of educatio... more In this article the authors start with a description of the learner-centered paradigm of education. The key tenets of the paradigm are outlined as: Competency-based student progress, competency-based student assessment and records, personal learning plans, project-based learning, just-in-time instructional support, student as self-directed learner, and teacher as guide on the side. Toward this end, we explain the self-directed, project-based learning approach using an exemplary school: Minnesota New Country School. Due to new roles of teachers and students in this new paradigm, we discuss how learning technology can support those roles by providing various functions. The functions include four major functions (record-keeping, planning, instruction, and assessment) and several secondary functions (communication, general student data, school staff information, secure and sustainable administration). In the final section, we address the need for transforming schools’ physical spaces and exemplify a design of such spaces to best support the learner-centered paradigm of education.
Bookmarks Related papers MentionsView impact
Existing Intelligent Tutoring Systems (ITSs) are unable to track affective states of learners. In... more Existing Intelligent Tutoring Systems (ITSs) are unable to track affective states of learners. In this paper, we focus on the problem of emotional engagement, and propose to detect important affective states (i.e., 'Satisfied', 'Bored', and 'Confused') of a learner in real time. We collected 210 hours of data from 20 students through authentic classroom pilots. The data included information from two modalities: (1) appearance which is collected from the camera, and (2) context-performance that is derived from the content platform. In this paper, data from nine students who attended the learning sessions twice a week are analyzed. We trained separate classifiers for different modalities (appearance and context-performance), and for different types of learning sections (instructional and assessment). The results show that different sources of information are generically better representatives of engagement at different sections: For instructional sections, generic appearance classifier yields higher accuracy (55.79%); whereas context-performance classifier is more accurate for assessment sections (63.41%). Moreover, the results indicate that expression of engagement is person-specific through both of these sources, and personalized engagement models perform more accurately: When person-specific data are added to the training set, on instructional sections, 85.44% and 96.13% accuracies are achieved for appearance and context-performance, respectively. For assessment sections, the accuracies are 75.25% (appearance) and 90.24% (context-performance). When only person-specific data are employed during training, similar accuracies are achieved even with very limited data.
Bookmarks Related papers MentionsView impact
Segmenting principle claiming that students learn better when a multimedia presentation is divide... more Segmenting principle claiming that students learn better when a multimedia presentation is divided into smaller segments than as a continuous unit (Mayer, 2001) was investigated in this study. The purpose of this study was to investigate the effects of having too many segments in a multimedia presentation on students' learning. In this study, two different multimedia presentations were used. One of the presentations was a segmented presentation (SP) of 10 slides in which the subject was the climate types. The other one was an over-segmented presentation (OSP) of 19 slides. Both classes were taken pretest about the subject and one week later same test was given as posttest. As a result of this, there was no significant difference (t =-0.758, p = 0.454 > .05) on retention between SP and OSP groups. In other words, oversegmenting the slides in a multimedia environment does not have a negative effect on students' understanding. Özet Bu araştırmada bir çoklu ortam sunumu sürekli devam eden bir bütün şeklinde olmasından ziyade küçük bölümlere ayrıldığındaöğrencilerin daha iyi öğrendiğini iddia eden parçalara bölme ilkesiaraştırılmıştır. Araştırmanın amacı, sunumunçok fazla bölümlere ayrılmasının öğrencilerin öğrenmesi üzerindeki etkisinin incelenmesidir. Bu araştırmada iki farklı çoklu ortam sunumu kullanılmıştır. Sunumlardan biri 10 slayttan oluşan iklim tipleriyle ilgili parçalara bölünmüş bir sunumdur(SP). Diğer sunum ise yine aynı konuyla ilgili olup 19 slayttan oluşan fazla parçaya bölünmüş bir sunumdur(OSP). Her iki sınıfa da önce konuyla ilgili bir ön test uygulanmıştır, bir hafta sonrasında da sunumlar gösterilip ardından gruplara aynı test son test olarak uygulanmıştır. Bu araştırmanın sonucunda SP ve OSP gruplarının kalıcılık testleri sonuçları arasında anlamlı bir farklılığa rastlanmamıştır(t =-0.758, p = 0.454 > .05). Başka bir deyişle, bir sunumu çok fazla parçaya bölmenin öğrencilerin anlamaları üzerinde olumsuz bir etkisi ortaya çıkmamıştır. Anahtar Kelimeler: Çoklu ortam, sunum, parçalara bölme ilkesi, fazla parçaya bölünmüş çoklu ortam sunumları.
Bookmarks Related papers MentionsView impact
Education is ripe for transformation given the world's social and economic development and needs ... more Education is ripe for transformation given the world's social and economic development and needs for new skills. Education systems around the world are looking for ways to leverage as well as contribute to the growing climate of innovation and creativity all while keeping up with changes in society. Advancing education reform particularly in countries with a growing youth population will encourage responsible citizenship, develop their human capacity and support their transition into a knowledge based economy. In recent years, education has been transformed from a teacher-centered approach to a learner-centered one. Innovative applications in education such as digital games, virtual reality, and robotics have gained importance to support learners and teachers with this transformation and to develop 21 st century skills of learners such as problem solving, collaboration, creativity, and critical thinking. Information-age society and the workforce expect students to possess these skills. In addition, studies show that ICT can help increase student engagement, motivation and attendance. Educators ought to bring innovative applications to classrooms and benefit from their motivational power to prepare students for their future. The potential for eLearning to improve performance on core subjects and foster the development of 21 st century skills depends on the schools ability to model student-centered, highly personalized learning environments supported by innovative teaching practice, administrative support and transformative policies to enable these learning strategies. Intel's Education Transformation Model is a proven, holistic
Bookmarks Related papers MentionsView impact
In our longitudinal research, we have been working towards an adaptive learning system automatica... more In our longitudinal research, we have been working towards an adaptive learning system automatically detecting student engagement as a higher-order user state in real-time. The labeled data necessary for supervised learning can be obtained through labeling conducted by human experts. Using multiple labelers to label collected data and obtaining agreement among different labelers on same samples of data is critical to train final engagement model accurately. Addressing these challenges, we developed a rigorous labeling process (HELP) specific to educational context with multi-faceted labels and multiple expert labelers. HELP has three distinct stages: (1) Pre-Labeling, including planning, labeler recruitment, training, and evaluation steps; (2) Labeling, involving actual labeling conducted by multiple labelers, and related steps for formative assessment of their performance; and (3) Post-Labeling, generating final labels and agreement measures through processing multiple decisions. In this paper, we outline proposed methods in HELP and describe the developed labeling tool.
Bookmarks Related papers MentionsView impact
Conference Presentations by Sinem Emine Mete
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Uploads
Papers by Sinem Emine Mete
Conference Presentations by Sinem Emine Mete