Benjamin Fankhauser
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Benjamin Fankhauser has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Kaspar Riesen
University of Bern
8 shared publications - Vidushi Bigler
Bern University of Applied Sciences
8 shared publications - Bangchao Deng
University of Macau, Macao SAR, China
1 shared publication - Bingqing Qu
BNU-HKBU United International College, Zhuhai, China
1 shared publication - Dingqi Yang
State Key Laboratory of Internet of Things for Smart City, Department of Computer and Information Science, University of Macau, Macau, China
1 shared publication - Linlin Jia
University of Bern
1 shared publication
Latest Publications
- 1 October 2026
- Published by Springer Nature
- p. 318-332
Embedding Techniques for Modeling Ambiguous Time Series in River Water Networks
- 1 October 2026
- Published by Springer Nature
- p. 318-332
Embedding Techniques for Modeling Ambiguous Time Series in River Water Networks
- Open Access
- 2 September 2026
- Published by Elsevier in SSRN Electronic Journal
How Graph Design Shapes Spatio-Temporal River Water Temperature Modeling
River water temperature modeling is important for environmental monitoring and water management, especially under climate change. Recent approaches use graph-based representations to capture spatial dependencies between monitoring stations, typically based on hydrological flow. However, it remains unclear which graph structures are most suitable for representing river systems. This paper systematically evaluates the impact of various graph construction. We compare Natural Flow Graphs, Relative Neighbor Graphs, Gabriel Graphs, Delaunay Triangulation Graphs, and k-nearest Neighbor Graphs across three real-world datasets using a spatio-temporal graph neural network for air-to-water modeling. Results show that hydrological flow graphs provide a strong baseline but are not consistently optimal. Proximity-based structures often achieve lower prediction errors, with Gabriel, Delaunay, and moderately dense k-nearest Neighbor Graphs performing best. An ablation study replacing graph neural networks with multi-head self-attention achieves competitive performance but does not outperform graph-based approaches. Overall, graph construction substantially influences predictive performance.
- 2 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessHow Graph Design Shapes Spatio-Temporal River Water Temperature Modeling
River water temperature modeling is important for environmental monitoring and water management, especially under climate change. Recent approaches use graph-based representations to capture spatial dependencies between monitoring stations, typically based on hydrological flow. However, it remains unclear which graph structures are most suitable for representing river systems. This paper systematically evaluates the impact of various graph construction. We compare Natural Flow Graphs, Relative Neighbor Graphs, Gabriel Graphs, Delaunay Triangulation Graphs, and k-nearest Neighbor Graphs across three real-world datasets using a spatio-temporal graph neural network for air-to-water modeling. Results show that hydrological flow graphs provide a strong baseline but are not consistently optimal. Proximity-based structures often achieve lower prediction errors, with Gabriel, Delaunay, and moderately dense k-nearest Neighbor Graphs performing best. An ablation study replacing graph neural networks with multi-head self-attention achieves competitive performance but does not outperform graph-based approaches. Overall, graph construction substantially influences predictive performance.
- Open Access
- 1 September 2026
- Published by Elsevier in SSRN Electronic Journal
Benchmarking Graphs and Embeddings for River Temperature Prediction
In climate research, modeling river water temperature is a key pattern recognition task for detecting trends and predicting change. As a geospatial time series problem, its predictive performance is highly sensitive to sensor choice and temporal coverage, limiting comparability across studies. As the first major contribution, we curate three river water temperature datasets with a well-defined experimental setup specifying sensor choice and temporal partitions to enable fair and robust comparisons across diverse scenarios. Building on this foundation, our second major contribution examines the role of learnable embeddings and structural information -- two recently introduced concepts in river water temperature modeling. To this end, we evaluate four models representing all combinations of these components: LSTM, Graphlet, LSTM-E, and ST-GNN (a spatio-temporal graph neural network). Our experiments corroborate the predictive performance reported in prior studies, while revealing substantial variability across water stations. Crucially, we provide statistical evidence that structural information consistently enhances baseline models across all datasets. Moreover, the marked reduction in learnable parameters -- a defining feature of embedding-based methods -- demonstrates their efficiency and underscores their potential for future river water temperature modeling architectures.
- 1 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessBenchmarking Graphs and Embeddings for River Temperature Prediction
In climate research, modeling river water temperature is a key pattern recognition task for detecting trends and predicting change. As a geospatial time series problem, its predictive performance is highly sensitive to sensor choice and temporal coverage, limiting comparability across studies. As the first major contribution, we curate three river water temperature datasets with a well-defined experimental setup specifying sensor choice and temporal partitions to enable fair and robust comparisons across diverse scenarios. Building on this foundation, our second major contribution examines the role of learnable embeddings and structural information -- two recently introduced concepts in river water temperature modeling. To this end, we evaluate four models representing all combinations of these components: LSTM, Graphlet, LSTM-E, and ST-GNN (a spatio-temporal graph neural network). Our experiments corroborate the predictive performance reported in prior studies, while revealing substantial variability across water stations. Crucially, we provide statistical evidence that structural information consistently enhances baseline models across all datasets. Moreover, the marked reduction in learnable parameters -- a defining feature of embedding-based methods -- demonstrates their efficiency and underscores their potential for future river water temperature modeling architectures.
- 3 August 2026
- Published by Springer Nature
- p. 250-264
Benchmarking Transformers on Spatio-Temporal River Water Temperature Modeling
- 3 August 2026
- Published by Springer Nature
- p. 250-264
Benchmarking Transformers on Spatio-Temporal River Water Temperature Modeling
- 8 June 2025
- Published by Springer Nature
- p. 203-213
Exploring a Graph Regression Problem in River Networks
- 8 June 2025
- Published by Springer Nature
- p. 203-213
Exploring a Graph Regression Problem in River Networks
- 31 January 2025
- Published by Springer Nature
- p. 31-40
Spatio-Temporal Graph Neural Networks for Water Temperature Modeling
- 31 January 2025
- Published by Springer Nature
- p. 31-40
Spatio-Temporal Graph Neural Networks for Water Temperature Modeling
- 19 September 2024
- Published by Springer Nature
- p. 283-294
Leveraging LSTM Embeddings for River Water Temperature Modeling
- 19 September 2024
- Published by Springer Nature
- p. 283-294
Leveraging LSTM Embeddings for River Water Temperature Modeling
- Open Access
- 30 September 2023
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 14 (5) , 1-24
Robust Location Prediction over Sparse Spatiotemporal Trajectory Data: Flashback to the Right Moment!
As a fundamental problem in human mobility modeling, location prediction forecasts a user’s next location based on historical user mobility trajectories. Recurrent neural networks (RNNs) have been widely used to capture sequential patterns of user visited locations for solving location prediction problems. Due to the sparse nature of real-world user mobility trajectories, existing techniques strive to improve RNNs by incorporating spatiotemporal contexts into the recurrent hidden state passing process of RNNs using context-parameterized transition matrices or gates. However, such a scheme mismatches universal spatiotemporal mobility laws and thus cannot fully benefit from rich spatiotemporal contexts encoded in user mobility trajectories. Against this background, we propose Flashback++, a general RNN architecture designed for modeling sparse user mobility trajectories. It not only leverages rich spatiotemporal contexts to search past hidden states with high predictive power but also learns to optimally combine them via a hidden state re-weighting mechanism, which significantly improves the robustness of the models against different settings and datasets. Our extensive evaluation compares Flashback++ against a sizable collection of state-of-the-art techniques on two real-world location-based social networks datasets and one on-campus mobility dataset. Results show that Flashback++ not only consistently and significantly outperforms all baseline techniques by 20.56% to 44.36% but also achieves better robustness of location prediction performance against different model settings (different RNN architectures and numbers of hidden states to flash back), different levels of trajectory sparsity, and different train-testing splitting ratios than baselines, yielding an improvement of 31.05% to 94.60%.
- 30 September 2023
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Intelligent Systems and Technology
- Vol. 14 (5) , 1-24
Open AccessRobust Location Prediction over Sparse Spatiotemporal Trajectory Data: Flashback to the Right Moment!
As a fundamental problem in human mobility modeling, location prediction forecasts a user’s next location based on historical user mobility trajectories. Recurrent neural networks (RNNs) have been widely used to capture sequential patterns of user visited locations for solving location prediction problems. Due to the sparse nature of real-world user mobility trajectories, existing techniques strive to improve RNNs by incorporating spatiotemporal contexts into the recurrent hidden state passing process of RNNs using context-parameterized transition matrices or gates. However, such a scheme mismatches universal spatiotemporal mobility laws and thus cannot fully benefit from rich spatiotemporal contexts encoded in user mobility trajectories. Against this background, we propose Flashback++, a general RNN architecture designed for modeling sparse user mobility trajectories. It not only leverages rich spatiotemporal contexts to search past hidden states with high predictive power but also learns to optimally combine them via a hidden state re-weighting mechanism, which significantly improves the robustness of the models against different settings and datasets. Our extensive evaluation compares Flashback++ against a sizable collection of state-of-the-art techniques on two real-world location-based social networks datasets and one on-campus mobility dataset. Results show that Flashback++ not only consistently and significantly outperforms all baseline techniques by 20.56% to 44.36% but also achieves better robustness of location prediction performance against different model settings (different RNN architectures and numbers of hidden states to flash back), different levels of trajectory sparsity, and different train-testing splitting ratios than baselines, yielding an improvement of 31.05% to 94.60%.
- 24 August 2023
- Published by Springer Nature
- p. 172-181
Graph-Based Deep Learning on the Swiss River Network
- 24 August 2023
- Published by Springer Nature
- p. 172-181
Graph-Based Deep Learning on the Swiss River Network