Abstract:The rising popularity of online social network services has attracted lots of research on mining social media data, especially on mining social events. Social event detection, due to its wide applications, has now become a trivial task. State-of-the-art approaches exploiting Graph Neural Networks (GNNs) usually follow a two-step strategy: 1) constructing text graphs based on various views (\textit{co-user}, \textit{co-entities} and \textit{co-hashtags}); and 2) learning a unified text representation by a specific GNN model. Generally, the results heavily rely on the quality of the constructed graphs and the specific message passing scheme. However, existing methods have deficiencies in both aspects: 1) They fail to recognize the noisy information induced by unreliable views. 2) Temporal information which works as a vital indicator of events is neglected in most works. To this end, we propose ETGNN, a novel Evidential Temporal-aware Graph Neural Network. Specifically, we construct view-specific graphs whose nodes are the texts and edges are determined by several types of shared elements respectively. To incorporate temporal information into the message passing scheme, we introduce a novel temporal-aware aggregator which assigns weights to neighbours according to an adaptive time exponential decay formula. Considering the view-specific uncertainty, the representations of all views are converted into mass functions through evidential deep learning (EDL) neural networks, and further combined via Dempster-Shafer theory (DST) to make the final detection. Experimental results on three real-world datasets demonstrate the effectiveness of ETGNN in accuracy, reliability and robustness in social event detection.

Event Detection As Graph Parsing

Event Detection Using a Self-Constructed Dependency and Graph Convolution Network

EventGraph: Event Extraction as Semantic Graph Parsing

Event Detection with Dual Relational Graph Attention Networks.

Event Detection with Multi-Order Edge-Aware Graph Convolution Networks

JSEEGraph: Joint Structured Event Extraction as Graph Parsing

Event Detection With Dynamic Word-Trigger-Argument Graph Neural Networks

Edge-Enhanced Graph Convolution Networks for Event Detection with Syntactic Relation

A Graph Convolutional Network with Adaptive Graph Generation and Channel Selection for Event Detection.

Syntax-based dynamic latent graph for event relation extraction

A Multi-Graph Representation for Event Extraction

Event detection from text using path-aware graph convolutional network

Syntax grounded graph convolutional network for joint entity and event extraction

ES4ED: An Event Detection Model Based on Event Sentence Pre-Determination

Event Detection on Dynamic Graphs

GTN-ED: Event Detection Using Graph Transformer Networks

Extracting Temporal Event Relation with Syntax-guided Graph Transformer

A Multi-Channel Hierarchical Graph Attention Network for Open Event Extraction

Graph-based Asynchronous Event Processing for Rapid Object Recognition

Evidential Temporal-aware Graph-based Social Event Detection via Dempster-Shafer Theory

Trigger-Argument based Explanation for Event Detection