The first DARUMA project on-site workshop was held in Budapest between Oct. 10 and Oct. 11, 2022. Project participants from KU (Japan), TUM (Germany), BME (Hungary) and UCM (Spain), and guests from the University of Luxembourg, got together to report the processes of the respective work packages in charge, and more importantly, to discuss next steps and future collaborations. According to the workshop agenda, Dr. Jan-Dirk Schmöcker from KU opened the workshop with an overview of the DARUMA project and suggested some important goals that should be achieved through this workshop.
Then, Qinglong Lu from TUM presented the process of W2 (Data Assembly), for which a GitHub repository has been created to share codes for preprocessing the datasets in common among the three study cities (i.e., Kyoto, Madrid, and Budapest). Lu also presented their works in OD estimation using LBSN data, and simulation-based policy assessment framework to quantify the traffic externalitie.
In the afternoon, Dr. Gustavo Romanillos Arroyo from UCM presented the process of WP3 (Data Fusion) and the method to conduct comparative analysis. Enrique Santiago Iglesias shared their findings from the GPT data of nightlife activity venues. They also organized three sub-workshops with the aim of exchanging research ideas and promoting future collaboration. The first day closed with a summary of collaboration topics achieved within the three sub-workshops.
On the second day of the workshop, Dr. Tamás Tettamanti from BME presented their discoveries in the relationship between venue popularity and traffic flow/speed, which stirred up a heated discussion about the function of GPT data and its futuristic possibility in describing urban traffic.
After a short break, the guest scholars from the University of Luxembourg, Dr. Richard Connors, Piergiorgio Vitello, and Nicola Schwemmle, presented their works on GPT data, particularly on the correlation between public transport station popularity and demand (entrances and exits). A specific “signature” index was designed to measure the characteristics of public transport stations based on the popularity trends observed.
In the afternoon, Dr. Jan-Dirk Schmöcker, Jiannan Dai, and Dr. Wenzhe Sun from KU, presented their works on using Twitter data to analyse the moods of people connected to travel during COVID-19, evidence of the long-term effect of COVID-19 discovered from GPT data, and the application of the GCN model in population forecast. Further, based on Sun’s presentation, all participants discussed the potential application of the GCN framework on the DARUMA project.
Finally, Dr. Jan-Dirk Schmöcker closed this workshop with a review of the goals to achieve listed at the beginning of the workshop and expressed the expectations of the DARUMA project. We would like to thank BME, especially Dr. Domokos Esztergár-Kiss and Dr. Tamás Tettamanti, for organising this workshop, and thank all project partners from Japan, Germany and Madrid for participating in discussions about the topics in concern. We would also like to thank our peers from the University of Luxembourg for joining the workshop and their contributions to this workshop with their valuable experience in GPT data analysis.