At this moment, there are two main research threads in our lab: educational human-data interaction (EDU-HDI) and socio-spatial human-data interaction (SS-HDI).

Here are 5 examples of our current research:

1. Analysis of Learning Behaviors (EDU-HDI)

Improving teaching and learning practices can be extremely difficult without proper understanding of relevant learning behaviors.

We design, develop, and evaluate HDI methods, systems and tools that support teachers and students gain proper understanding of learning-relevant behaviors by employing data mining and information visualization techniques. This work is based on our collaboration with Learning Analytics Center.

  • Fumiya Okubo, Masanori Yamada, Misato Oi, Atsushi Shimada, Yuta Taniguchi and Shin’ichi Konomi (2019). Learning Support Systems Based on Cohesive Learning Analytics. Emerging Trends in Learning Analytics, Brill, ISBN: 978-90-04-39661-6

  • Yuta Taniguchi, Atsushi Shimada, Shin’ichi Konomi (2019). Investigating Error Resolution Processes in C Programming Exercise Courses. Proceedings of the 12th International Conference on Educational Data Mining.

2. Learning Analytics for All (EDU-HDI)

An increasing number of communities are super-aged. However, most of the existing digital learning-support environments cannot easily support older adults in recurrent and lifelong learning contexts.

We design, develop, and evaluate HDI methods, systems and tools that support older adults by employing user-centered approaches and sensor-based analytics to make learning tools and materials easier to use and more effective for people over the age of 65. See also: the website of a relevant project.

  • Learning Analytics for All: Opportunities and Challenges, The 7th Asian Workshop on Smart Sensor Systems, Munakata, March 24-26, 2019 [Keynote Talk]

  • Min Lu, Kaoru Tamura, Shin’ichi Konomi (2019). An Elderly-Oriented User Interface Prototype Developed for Inclusive Learning Support Systems. Presented at the 2019 Annual International Conference on Education and Service Sciences (ICESS 2019), Wuhan China, September 20-23, 2019. [Best Presentation Award]

  • Shin’ichi Konomi, Kohei Hatano, Miyuki Inaba, Misato Oi, Tsuyoshi Okamoto, Fumiya Okubo, Atsushi Shimada, Jingyun Wang, Masanori Yamada, Yuki Yamada (2018). Towards Supporting Multigenerational Co-Creation and Social Activities: Extending Learning Analytics Platforms and Beyond. Proceedings of the 6th International Conference on Distributed, Ambient, and Pervasive Interactions (DAPI 2018), Held as Part of HCI International 2018, Las Vegas, NV, July 15-20, 2018, pp.82-91. Lecture Notes in Computer Science, Springer, Berlin/Heidelberg, 2018.

3. Recommender Systems in University Environments (EDU-HDI)

Existing course recommendation systems for online learning environments could not fully support students in university environments due to the inherent differences of the online and physically-based learning environments.

We design, develop, and evaluate HDI methods, systems and tools that recommend elective courses to university students by employing recommender algorithms and systems that are aware of relevant constraints and resources in university environments, thereby supporting students find the right courses to take.

  • Boxuan Ma, Yuta Taniguchi, Shin’ichi Konomi (2020). Course Recommendation for University Environment. Proceedings of the 13th International Conference on Educational Data Mining (EDM 2020), International Educational Data Mining Society (IEDMS), Worcester, pp. 460 – 466.

  • Boxuan Ma, Min Lu, Yuta Taniguchi and Shin’ichi Konomi (2020). Exploring the Design Space for Explainable Course Recommendation Systems in University Environments. Comanion Proceedings of the 10th International Conference on Learning Analytics & Knowledge (LAK20), pp. 492-499.

  • Boxuan Ma (2019). Design of an Elective Course Recommendation System for University Environment, Proceedings of the 12th International Conference on Educational Data Mining (EDM2019), Doctoral Consortium, Montreal, Canada, July 2-5, 2019.

4. Mobile and Situated Crowdsourcing (SS-HDI)

Data Science requires a surprisingly large amount of human labor collecting, labeling, preparing, and using data, which is a particularly critical issue when we are to deal with real-world data and services in specific spaces and situations. This necessiates development of efficient approaches to crowdsource data-related labor in different spaces and situations.

We design, develop, and evaluate HDI methods, systems and tools that support efficient mobile and situated crowdsourcing in public spaces and local communities to collect useful data without requiring too much human labor. We do this by employing crowdsourcing methods, algorithms, and tools that are aware of relevant contexts and models.

  • Samuli Hemminki, Keisuke Kuribayashi, Shin’ichi Konomi, Petteri Nurmi, Sasu Tarkoma (2019). Crowd Replication: Sensing-Assisted Quantification of Human Behavior in Public Spaces. ACM Transactions on Spatial Algorithms and Systems

  • Shin’ichi Konomi, Tomoyo Sasao, Simo Hosio, Kaoru Sezaki . Using Ambient WiFi Signals to Find Occupied and Vacant Houses in Local Communities. Journal of Ambient Intelligence and Humanized Computing, pp.1-11. Springer, Berlin/Heidelberg.

5. Contextual Reminders (SS-HDI)

Providing useful feedback to the users at the right time, at the right place, and in the right way may require knowledge and wisdom that machines are not good at acquiring. Although existing Data Science approaches tend to treat humans merely as data sources, realizing effective data-driven feedback mechanisms may necessiates an alternative perspective on humans as the participants and/or social partners rather than the monitored.

We design, develop, and evaluate HDI methods, systems and tools that can deliver useful safety-related contextual reminders to the members of a local community. We do this by involving locals at diffent levels of the system to elicit and integrate their local knowledge and wisdom into the system’s information space. This is based on participatory design, and employs an active crowdsourcing technique as well as IoT-based ‘invisible’ user interface design. We thus facilitate participation by non tech-savvy local residents in the design as well as the data collection process of contextual reminders.

  • Tomoyo Sasao, Shin’ichi Konomi, Vassilis Kostakos, Keisuke Kuribayashi, Jorge Goncalves. Community Reminder: Participatory Contextual Reminder Environments for Local Communities. International Journal of Human-Computer Studies, 102, pp. 41-53, Elsevier, Amsterdam, June 2017.

Ongoing Experiments