New paper in the Journal of Big Data: Where you go is who you are

Location data is extremely sensitive, revealing where we spend how much time, what we like to do in our free time, and how we spend our money. Our new study titled “Where you go is who you are – A study on machine learning based semantic privacy attacks” sets out to answer a critical question: Just how much “semantic information” – that is, detailed and potentially sensitive insights into personal behavior – can be extracted from someone’s location data using machine learning techniques, especially when this data might be inaccurate or incomplete?

To address this question, we conducted a study using a dataset of location visits from a social network, Foursquare, where the category of the location (e.g., “dining” / “sports” / “nightlife”) is known. We simulate the following scenario: An attacker only knows the geographic coordinates and time of a location visit, and trains a machine learning model to infer the place category. Crucially, the geographic coordiates can be imprecise due to GPS inaccuracies or protection measures. This is simulated by obfuscating the true geographic coordinates with varying radius.

Our experiments reveal a significant privacy risk even with imprecise location data. Although the accuracy of the attacker decreases exponentially with the obfuscation radius, there remains a high privacy loss even if the coordinates are perturbed by up to 200m. A certain risk remains from the temporal information alone. These findings highlight the privacy risks associated with the growing databases of tracking and spatial context data, urging policy-makers towards stricter regulations.

Check out our paper and source code!

New paper published in the Journal of Transport Geography – Car sharing demand prediction

Our paper titled “Spatially-aware station based car-sharing demand prediction” is now published open-access in the Journal of Transport Geography!

In this paper, we analyze long-term station-based car-sharing demand (i.e., the monthly number of reservations per station), and fit local and global models to the demand. Spatially-aware models and methods for interpretability improved our understanding of the effect of different features in varying locations. Our models can assist in planning new car sharing stations, an important avenue towards more sustainable transportation.

This work is part of the Vehicle-to-grid for Car Sharing project.

New paper in Transportation Research Part D – Time flexibility of car sharing users

Our paper titled “Vehicle-to-grid and car sharing: Willingness for flexibility in reservation times in Switzerland” presents the results of a stated preference survey conducted with 777 Mobility car sharing users in Switzerland. This study was part of our Vehicle-to-grid for Car Sharing project, funded by the Swiss Federal Office of Energy, where we aim to analyze the potential of integrating V2G in car sharing services. Understanding the time flexibility of car sharing users is crucial for designing dynamic pricing strategies, for example with the goal to incentivize users to shift their reservation times and thereby to increase the flexibility for V2G.
We found the value of time to be 31CHF/h on average, where older adults, lower income groups and employed adults tend to have lower flexibility. For more details, see our paper published in Transportation Research Part D.

New JTRG paper online – Travel mode detection

Our new paper entitled “Evaluating geospatial context information for travel mode detection” was accepted at Journal of Transport Geography and is now available (open-access!) online.

How much does geospatial context information contribute to travel mode detection?

Our latest study reveals that geospatial network features, such as distance to the road network, are more critical than motion features, such as speed and acceleration, when classifying an extensive list of travel modes. Still, most land-use and land-cover features barely contribute to the task. The results are based on our extensive context representation reviews and the proposed analytical pipeline to assess the contribution of geospatial context information based on a random forest model and the SHapley Additive exPlanation (SHAP) method.

The study provides valuable guidance for feature selection, effective feature design, and building efficient travel mode detection models.

Check out the paper online and the corresponding code on Github!

New TR_C paper online – Context-aware next location prediction

Our new paper entitled “Context-aware multi-head self-attentional neural network model for next location prediction” was accepted at Transportation Research Part C: Emerging Technologies and is now available (open-access!) online.

We present a multi-head self-attentional (MHSA) neural network that integrates location features, temporal features, and functional land use contexts for next location prediction. This comprehensive approach effectively captures movement-related spatio-temporal information, leading to state-of-the-art performance on GNSS mobility datasets.

Our analysis demonstrates that training the model on population data yields superior results by learning from collective movement patterns, surpassing the capabilities of individual-level models. Moreover, we emphasize the significance of recent past movements and weekly periodicity, showing that learning from a subset of historical mobility is sufficient to obtain an accurate location prediction result.

The proposed model represents a pivotal advancement in accurate and interpretable individual mobility prediction, and can be readily applied in downstream applications, including planning on-demand transport services, implementing mobility incentives, and suggesting alternative mobility options.

Check out the paper online and the corresponding code on Github!

We invite you to join our workshop on Reproducibility in Tracking Data Analysis and Mobility Research at ACM SIGSPATIAL

This year at the ACM SIGSPATIAL conference, we are hosting a workshop on Reproducibility in Tracking Data Analysis and Mobility Research (https://github.com/mie-lab/reprotrack)!

Considering the fast methodological advances in spatial data science, the topic of reproducibility is more important than ever before. To foster common standards and transparency, we aim to bring researchers together in this session to discuss challenges and future pathways for reproducible spatial data science, with a focus on mobility data. The workshop is planned as a particularly interactive session, including a hands-on tutorial on tracking data preprocessing where you can bring your own data.

Please sign up here if you plan to attend the workshop. We hope to see you there on Monday, November 13th, in Hamburg!

New IJGIS Paper is online – using context data to improve traffic forecasting

Our new paper entitled “Incorporating multimodal context information into traffic speed forecasting through graph deep learning” is now online at IJGIS.

In this work, we propose a multimodal context-based graph convolutional neural network (MCGCN) model to fuse context data into traffic speed prediction, including spatial and temporal contexts. The proposed model comprises three modules, i.e., (a) hierarchical spatial embedding to learn spatial representations by organizing spatial contexts from different dimensions, (b) multivariate temporal modeling to learn temporal representations by capturing dependencies of multivariate temporal contexts and (c) attention-based multimodal fusion to integrate traffic speed with the spatial and temporal context representations for multi-step speed prediction. We conduct extensive experiments in Singapore. Compared to the baseline model (STGCN), our results demonstrate the importance of multimodal contexts with the mean-absolute-error improvement of 0.29 km/h, 0.45 km/h and 0.89 km/h in 30-min, 60-min and 120-min speed prediction, respectively. We also explore how different contexts affect traffic speed forecasting, providing references for stakeholders to understand the relationship between context information and transportation systems. Check out the open-access paper online!

Paper published in the Journal of LBS

Our paper titled “Influence of tracking duration on the privacy of individual mobility graphs” was published by the Journal of Location Based Services! In this work, we use a GPS tracking dataset and analyze how the tracking duration affects the risk for users to be re-identified; i.e., by matching to previously stored tracking data. It is well known that the tracking data of a user is quite unique and can be matched to stored data easily; however, we study the risk of a representation of tracking data that is already privatised, namely so-called location graphs. Location graphs do not reveal the geocoordinates or time stamps of the places that a user visited, but just the topology of the mobility behaviour. Nevertheless, users can be re-identified with a top-1 accuracy of up to 20%, and the re-identification risk strongly depends on the tracking duration of the user, as well as the duration of the stored data (pool), as shown in the figure below. Check out our paper and code for more information!