How can algorithms and mobile data be the future of statistics for tourism and mobility?
Motion Analytica, in collaboration with ISTAT and Vodafone Italia, helped develop an experimental project aimed at investigating some key issues related to tourism and mobility through the use of big data.
The details of the case study on mobility are described here. For the case study on tourism click here.
ISTAT & VODAFONE ITALY
This project has been presented at the 14th National Statistics Conference (Rome, Nov. 30-Dec. 1, 2021 –https://www.istat.it/it/archivio/261537) which is the main research meeting of the National Statistical System where data producers and users are given the opportunity to discuss the role and the future developments of official statistics.
OFFICIAL STATISTICS DATA
Provided by ISTAT.
ANONYMIZED DATA FROM MOBILE PHONE
Derived from traffic on phone cells in the analysed areas and provided by Vodafone italy.
The project involved the analysis of mobility of the whole territory of Lombardy, and in particular the province of Milan. The collaboration between ISTAT and Vodafone Italy made it possible to define an innovative data-processing framework and to test a new census strategy for mobility through the integration of new information sources.
The main goal was to create an Origin/Destination (O/D) matrix using anonymized mobile data to understand how people move during the week, as well as provide insights into their demographic profile and purpose of their travel.
The analysis covered a period of 3 months and several hundreds of areas, according to two main aspects:
Analysis of travel and commuting
The municipalities were aggregated into four groups based on the number of inhabitants, and all users who commuted between them.
All the data from telecom sources regarding travel and commuting in the Lombardy were compared with official ISTAT sources in order to assess the potential of the methodology.
The analytical approach of the project allowed for the detailed verification of several aspects related to people and their mobility, such as, for example, the comparison between inhabitants defined on the basis of telephone data and registry residents; the revision of commuting definitions; and the creation of O/D matrices with varying degrees of detail and fast updates.
This eventually led to reflect and discuss on the defining paradigms of information and their changes between what is explicitly collected through questionnaires and what is derived from statistical inference on big data.
Finally, the methodology allowed us to hypothesize the creation of innovative statistical products that can take advantage of the timeliness of the source not only for the creation of O/D matrices but also for estimates related, for example, to the identification of the means used for travel.
You can view the poster presented at the conference containing full details here: