Preaload Image
Back

Description

DiversityOne is one of the largest and most geographically diverse datasets for studying everyday life behavior of college students through smartphone sensors and self-reports.

Key statistics

Questionnaires from 18K+ participants

Intensive longitudinal survey of 4 weeks

Data from both the Global North and the Global South

Passive smartphone sensor data and self-reports from 782 participants across eight universities in eight different countries

Raw sensor data from 26 modalities

Possible research directions, but not limited to:

  • developing machine learning models across multiple domains, such as behavior recognition
  • multimodal time-series modeling
  • model generalization and domain adaptation across countries
  • social practices understanding
  • explore the diversity of daily activities in various cultural contexts
  • cross-cultural studies
DiversityOne Paper
Request Datasets
Dataset Catalog

Dataset Description

The DiversityOne dataset was collected through a collaborative effort involving universities across eight different countries, encompassing both the Global North and Global South. This extensive data collection includes intensive longitudinal surveys and passive smartphone sensor data from diverse cultural contexts, providing a unique resource for cross-cultural studies and behavioral research.

Countries and Universities

Flag_of_China

China

Jilin University (JLU)
Flag_of_Denmark.svg

Denmark

Aalborg University (AAU)
Flag_of_India

India

Amrita Vishwa Vidyapeetham (AMRITA)
Flag_of_Italy

Italy

University of Trento (UNITN)
Flag_of_Mexico

Mexico

Instituto Potosino de Investigación Científica y Tecnológica (IPICYT)
Flag_of_Mongolia.svg

Mongolia

National University of Mongolia (NUM)
Paraguay_flag

Paraguay

Universidad Católica “Nuestra Señora de la Asunción” (UC)
Flag_of_the_United_Kingdom

United Kingdom

London School of Economics and Political Science (LSE)
The overall data collection protocol is presented in the following figure:

 

Figure 1: Data collection protocol

 

  • 1st Questionnaire: it was sent to all students of the participating universities. It implements standard scales for measuring personality traits (Big Five Inventory) and attitudes toward values (Basic Values Survey). See the list on the data catalog website.
  • 2nd Questionnaire: participants joining the smartphone data collection answered questions about social practices and from Jungian Scale for Personality Types and Human Values Survey scales.
  • 1st Data collection phase: the intensive longitudinal survey observes human behaviors over time. Participants installed the ILog app on their smartphone. The app collects self-reports (every 30 minutes and additional morning and evening questions) and passive sensor data.
  • 2nd Data collection phase: as in the 1st phase, but self-reports are asked every hour.
  • 3rd Questionnaire: it collects feedback about the data collection and the Multiple Intelligence Scale.

The collected sensors are repoted in the following table.

List of sensors. The type column reports HW for hardware sensors, and SW for software sensors. The dataset paper describes each sensor.
Category No Type Name Frequency
Connectivity 1 SW Bluetooth Devices Once every minute
2 SW Cellular network info Once every minute
3 SW WIFI Network Connected to On change
4 SW WIFI Networks Available Once every minute
Environment 5 HW Light Up to 10 samples per second
6 HW Pressure Up to 10 samples per second
Motion 7 HW Accelerometer Up to 10 samples per second
8 HW Gyroscope Up to 10 samples per second
9 SW Movement Activity Once every 30 seconds
10 SW Step Counter Up to 10 samples per second
11 SW Step Detection On change
Position 12 HW Location Once every minute
13 HW Magnetic Field Up to 10 samples per second
14 SW Proximity Up to 10 samples per second
App usage 15 SW Headset Status [ON/OFF] On change
16 SW Music Playback On change
17 SW Notifications On change
18 SW Running Applications Once every 5 seconds
Device usage 19 SW Airplane Mode [ON/OFF] On change
20 SW Battery Charge [ON/OFF] On change
21 SW Battery Level On change
22 SW Doze Mode [ON/OFF] On change
23 SW Ring mode [Silent/Normal] On change
24 SW Touch event On change
25 SW Screen Status [ON/OFF] On change
26 SW User Presence On change
 
Table 1: Number of participants for each of the three questionnaires and data collection. “iLog signed” and “iLog data” columns show the total number of students who logged into the iLog app and those who actively contributed with data, respectively. Additionally, demographic information of the participants involved in the iLog data collection is reported.
Country Acronym 1st QU 2nd QU 3rd QU iLog signed iLog data μ Age (σ) % women
China JLU 1,007 69 41 54 44 19.4 (2.2) 61
Denmark AAU 412 16 15 27 20 28.2 (6.3) 58
India AMRITA 4,183 141 38 64 45 21.4 (2.9) 29
Italy UNITN 5,692 287 215 263 241 22.1 (3.2) 56
Mexico IPICYT 40 9 11 21 21 25.2 (4.1) 33
Mongolia NUM 3,972 214 152 231 201 20.0 (3.1) 65
Paraguay UC 1,342 33 25 43 28 23.3 (5.1) 61
UK LSE 1,980 143 45 88 66 24.6 (5.0) 66
Total 18,628 912 542 782 666

How to get the Dataset

1. Explore the Data Catalog – Browse the data catalog to identify the datasets relevant to your research.
2. Submit a Request – Visit the DataScientia Marketplace to submit your research proposal and request access to datasets. Fill out the request form, specifying your research purpose and selecting the desired datasets. 
3. Receive Approval & Download – Once your request is approved, you will receive an email with instructions on how to download the dataset.

Additional information can be found on the catalog. For any issue or question, please feel free to contact us at datadistribution.knowdive@unitn.it.

How to cite

@article{busso2025diversityone,
title = {{{DiversityOne}}: {{A Multi-Country Smartphone Sensor Dataset}} for {{Everyday Life Behavior Modeling}}},
author = {Busso, Matteo and Bontempelli, Andrea and Javier Malcotti, Leonardo and Meegahapola, Lakmal and Kun, Peter and Diwakar, Shyam and Nutakki, Chaitanya and Rodas Britez, Marcelo and Xu, Hao and Song, Donglei and Ruiz-Correa, Salvador and Mendoza-Lara, Andrea-Rebeca and Gaskell, George and Stares, Sally and Bidoglia, Miriam and Ganbold, Amarsanaa and Chagnaa, Altangerel and Cernuzzi, Luca and Hume, Alethia and Chenu-Abente, Ronald and Alia Asiku, Roy and Kayongo, Ivan and Gatica-Perez, Daniel and De Götzen, Amalia and Bison, Ivano and Giunchiglia, Fausto},
journal={Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies},
year={2025},
publisher={ACM New York, NY, USA}
}

Matteo Busso, Andrea Bontempelli, Leonardo Javier Malcotti, Lakmal Meegahapola, Peter Kun, Shyam Diwakar, Chaitanya Nutakki, Marcelo Rodas Britez,Hao Xu, Donglei Song, Salvador Ruiz-Correa, Andrea-Rebeca Mendoza-Lara, George Gaskell, Sally Stares, Miriam Bidoglia, Amarsanaa Ganbold, Altangerel Chagnaa, Luca Cernuzzi, Alethia Hume, Ronald Chenu-Abente, Roy Alia Asiku, Ivan Kayongo, Daniel Gatica-Perez, Amalia De Götzen, Ivano Bison, and Fausto Giunchiglia. (2025). DiversityOne: A Multi-Country Smartphone Sensor Dataset for Everyday Life Behavior Modeling. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies.

Studies based on DiversityOne

2024

  • Meegahapola et al. 2024. M3BAT: Unsupervised Domain Adaptation for Multimodal Mobile Sensing with Multi-Branch Adversarial Training. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.  Paper accessible here.
    Keywords: social context, mood
  • Mäder et al. 2024. Learning About Social Context from Smartphone Data: Generalization Across Countries and Daily Life Moments. In Proceedings of the CHI Conference on Human Factors in Computing Systems. Paper accessible here. 
    Keywords: social context

2023

  • Meegahapola et al. 2023. Generalization and personalization of mobile sensing-based mood inference models: an analysis of college students in eight countries. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies. Paper accessible here. 
    Keywords: mood
  • Assi et al. 2023. Complex daily activities, country-level diversity, and smartphone sensing: A study in Denmark, Italy, Mongolia, Paraguay, and UK. In Proceedings of the 2023 CHI conference on human factors in computing systems. Paper accessible here. 
    Keywords: daily activities
  • Kammoun et al. 2023. Understanding the Social Context of Eating with Multimodal Smartphone Sensing: The Role of Country Diversity. In Proceedings of the 25th International Conference on Multimodal Interaction. Paper accessible here. 
    Keywords: social context
  • Girardini et al. 2023. Adaptation of student behavioural routines during Covid-19: a multimodal approach. EPJ Data Science. Paper accessible here. 
    Keywords: daily routines

2022

  • Bouton-Bessac et al. 2022. Your Day in Your Pocket: Complex Activity Recognition from Smartphone Accelerometers. In International Conference on Pervasive Computing Technologies for Healthcare. Paper accessible here. 
    Keywords: daily activities

Related links

News

  • 2025-02-05: We are organizing an open challenge on the DiversityOne dataset at UbiComp 2025. Visit the workshop webpage for more details.
  • 2025-02-03: The dataset paper has been accepted on UbiComp/Imwut.

Contacts

Contact us for any questions or issues related to the datasets.

Acknowledgments

This research has received funding from the European Union’s Horizon 2020 FET Proactive project “WeNet – The Internet of us”, grant agreement No. 823783. We deeply thank all the volunteers across the world for their participation in the study. We thank the anonymous reviewers for their valuable feedback. We acknowledge the use of ChatGPT and Grammarly as tools for grammar refinement. University of Trento, Jilin University, National University of Mongolia and Universidad Católica “Nuestra Señora de la Asunción” have participated to the work described in this paper as part of the DataScientia initiative.

 

Related Items