BINTS

Bi-Modal Learning for Networked Time Series

1. Overview

Understanding human mobility patterns is a complex challenge that requires modeling both node features and node interactions within graph topologies across time. While previous methods have focused on either node features or interactions, the synergistic integration of these two modalities has proven difficult to achieve. In this paper, we propose BINTS (BI-modal learning for Networked Time Series), a pioneering bi-modal learning framework that employs soft contrastive learning along the temporal axis. BINTS captures modality similarities and temporal patterns by simultaneously learning from evolving node features and interactions, solving the limitations of single-modality approaches. To evaluate our method, we curate comprehensive multi-modal human mobility datasets spanning diverse locations and times. Our experimental results demonstrate that BINTS significantly outperforms existing forecasting models by capturing synergies across different data modalities. Overall, we establish BINTS as a powerful technique for holistically understanding and forecasting complex mobility dynamics.

2. Proposed Benchmark Datasets

Due to the big dataset size, we released it on the anonymous drive

• For research drive
• For benchmark drive

Domain	Dataset	#Node	Target Dim.	Total Period	Train Days	Test Days	Time Interval
Transportation	Daegu	85	7395	2022.01.01 ~ 2022.12.31	274	91	1 hour
	Busan	103	10815
	Seoul	233	54755
	NYC Taxi	10	120	2014.01.01 ~ 2023.12.31	2739	913	1 hour
Epidemic	COVID	16	272	2020.01.20 ~ 2023.08.31	990	330	1 day
	NYC COVID	5	30	2020.03.01 ~ 2023.12.31	1051	350	1 day

3. Requirements and Installations

• Node.js: 16.13.2+
• Anaconda 4 or Miniconda 3
• Python 3.11.5 (Recommend Anaconda)
• Ubuntu 18.04.6 LTS
• pytorch >= 2.1.2

4. Configuration

BINTS was implemented in Python 3.11.5.

• Edit main.py file to set experiment parameters (dataset, seq_length, gpu_id(e.g. 0,1,2,3,4,5), etc.)

python3 main.py

5. How to run

• Parameter options

--dataset: the name of dataset (string)
--seq_day: the size of a lookback window (integer)
--pred_day: the size of a prediction window (integer)

• At current directory which has all source codes, run main.py with parameters as follows.

- dataset: {busan, daegu, seoul, covid, nyc, nyc_covid}
- loop: {0, 1, 2, 3, 4}                       # seed for 5-fold evaluation.
- gpu_id: an integer gpu id.

e.g.) python3 main.py --gpu_id 1 --multi_gpu 1 --batch_size 8 --dataset nyc --seq_day 4 --pred_day 7 --khop 5