EasyTPP: Towards Open Benchmarking Temporal Point Processes
Siqiao Xue, Xiaoming Shi, Zhixuan Chu, Yan Wang, Hongyan Hao, Fan, Zhou, Caigao Jiang, Chen Pan, James Y. Zhang, Qingsong Wen, Jun Zhou,, Hongyuan Mei

TL;DR
EasyTPP is a comprehensive, open-source benchmark repository for temporal point processes, facilitating standardized evaluation, comparison, and reproducibility to accelerate research and application in event sequence modeling.
Contribution
It introduces the first centralized benchmark platform for TPPs, including datasets, evaluation tools, and model implementations, promoting reproducibility and collaborative research.
Findings
Provides a unified interface for datasets and models
Includes diverse evaluation programs for reproducibility
Offers implementations of popular neural TPPs and modular components
Abstract
Continuous-time event sequences play a vital role in real-world domains such as healthcare, finance, online shopping, social networks, and so on. To model such data, temporal point processes (TPPs) have emerged as the most natural and competitive models, making a significant impact in both academic and application communities. Despite the emergence of many powerful models in recent years, there hasn't been a central benchmark for these models and future research endeavors. This lack of standardization impedes researchers and practitioners from comparing methods and reproducing results, potentially slowing down progress in this field. In this paper, we present EasyTPP, the first central repository of research assets (e.g., data, models, evaluation programs, documentations) in the area of event sequence modeling. Our EasyTPP makes several unique contributions to this area: a unified…
Peer Reviews
Decision·ICLR 2024 poster
- Addresses a major need in the machine learning community interested in TPPs--the lack of a standardized benchmarking tool. Most researchers are piecing together implementations from other papers in order to make comparisons. Existing software packages, such as PoPPy are out of date and not maintained. The proposed EasyTPP package could fill a major need for the community. - Implements a comprehensive list of models and evaluation metrics, together with several representative data sets. - The a
- The authors compare against the classical Multivariate Hawkes Process (MHP) but don't specify what type of kernel they use (I assume exponential) or structure of the excitation matrix. Minor concerns: - Some typos, e.g. Multivariate Hakwes Process (MHP)
1. Good motivation
1. Lack of technique contribution 2. Unclear presentation
1. The most important contribution lies in providing a simple and standardized framework to allow users to apply different TPP models to any datasets, which promotes reproducible research in this field.
The main concern is contribution. Considering the quality of the ICLR community, although it is particularly urgent to propose a unified comparison and processing framework for point processes, the contribution is still limited.
Code & Models
Videos
Taxonomy
TopicsData Management and Algorithms · Constraint Satisfaction and Optimization · Graph Theory and Algorithms
MethodsLib
