Hierarchical and Incremental Structural Entropy Minimization for Unsupervised Social Event Detection
Authors: Yuwei Cao, Hao Peng, Zhengtao Yu, Philip S. Yu
AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments show that HISEvent consistently outperforms GNN-based methods and achieves the new SOTA for social event detection under both closedand open-set settings while being efficient and robust. |
| Researcher Affiliation | Academia | 1 Department of Computer Science, University of Illinois Chicago, Chicago, USA 2 School of Cyber Science and Technology, Beihang University, Beijing, China 3 Yunnan Key Laboratory of Artificial Intelligence, Kunming University of Science and Technology, Kunming, China |
| Pseudocode | Yes | Algorithm 1: Determine Es via incremental 1D SE minimization. Algorithm 2: Event detection via hierarchical 2D SE minimization. |
| Open Source Code | Yes | Our code is publicly available 1. 1https://github.com/SELGroup/HISEvent |
| Open Datasets | Yes | We experiment on two large, public Twitter datasets, i.e., Event2012 (Mc Minn, Moshfeghi, and Jose 2013), and Event2018 (Mazoyer et al. 2020). |
| Dataset Splits | No | We evaluate under both closedand open-set settings by adopting the data splits of Ren et al. 2022a and Cao et al. 2021. The paper refers to external data splits but does not provide specific percentages or counts for training, validation, or test sets within the provided text. |
| Hardware Specification | No | The paper does not provide specific hardware details used for running its experiments. |
| Software Dependencies | No | The paper mentions using SBERT and BERT but does not provide specific version numbers for software dependencies or libraries used in the implementation. |
| Experiment Setup | No | The paper mentions 'Implementation details are in Appendix' and discusses the hyperparameter 'n', but does not explicitly provide other specific hyperparameter values or comprehensive system-level training settings in the main text. |