From August 31 to September 3, 2026, we presented one oral presentation and one poster presentation from projects involving our lab at the 54th Annual Meeting of the Behaviormetric Society of Japan, held at Okayama University of Science.

The presentations explored how data can generate value through its use and combination from two different perspectives: value formation in data analytics communities and electricity demand forecasting using time series foundation models. The oral presentation analyzed the usage and co-use patterns of marketing-related datasets on Kaggle to investigate how their value is formed within a data analytics community. The poster presentation examined how time-derived covariates affect the accuracy and reliability of electricity demand forecasting using time series foundation models, particularly under difficult forecasting conditions.

Oral Presentation

Title: Value Formation Processes of Marketing Data in Data Analytics Communities

Presenters: Teruaki Hayashi, Takumi Shimizu, Hiroyasu Matsushima

Poster Presentation

Title: Electricity Demand Forecasting Using Time Series Foundation Models with Time-derived Covariates

Presenters: Han Taehee, Teruaki Hayashi