Complex Evolutional Pattern Learning for Temporal Knowledge Graph Reasoning

Li, Zixuan; Guan, Saiping; Jin, Xiaolong; Peng, Weihua; Lyu, Yajuan; Zhu, Yan; Bai, Linquan; Li, Wei; Guo, Jiafeng; Cheng, Xueqi

doi:10.18653/v1/2022.acl-short.32

Cited by 53 publications

(28 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…TITer [26] utilises reinforcement learning to model the temporal path and regards the process of temporal forecasting as the agent travelling on historical KG snapshots to search for the answer. CEN [7] notices the length‐diversity and time‐variability of evolutional patterns, and uses a length‐aware convolutional neural network to handle them. TiRGN [27] notes that the characteristics of historical facts include sequential, repetitive and cyclical patterns, so that it takes a comprehensive consideration of them.…”

Section: Related Workmentioning

confidence: 99%

Extrapolation over temporal knowledge graph via hyperbolic embedding

Jia

Wang

et al. 2023

CAAI Trans on Intel Tech

View full text Add to dashboard Cite

Predicting potential facts in the future, Temporal Knowledge Graph (TKG) extrapolation remains challenging because of the deep dependence between the temporal association and semantic patterns of facts. Intuitively, facts (events) that happened at different timestamps have different influences on future events, which can be attributed to a hierarchy among not only facts but also relevant entities. Therefore, it is crucial to pay more attention to important entities and events when forecasting the future. However, most existing methods focus on reasoning over temporally evolving facts or mining evolutional patterns from known facts, which may be affected by the diversity and variability of the evolution, and they might fail to attach importance to facts that matter. Hyperbolic geometry was proved to be effective in capturing hierarchical patterns among data, which is considered to be a solution for modelling hierarchical relations among facts. To this end, we propose ReTIN, a novel model integrating real-time influence of historical facts for TKG reasoning based on hyperbolic geometry, which provides low-dimensional embeddings to capture latent hierarchical structures and other rich semantic patterns of the existing TKG. Considering both real-time and global features of TKG boosts the adaptation of ReTIN to the ever-changing dynamics and inherent constraints. Extensive experiments on benchmarks demonstrate the superiority of ReTIN over various baselines. The ablation study further supports the value of exploiting temporal information.

show abstract

Section: Related Workmentioning

confidence: 99%

Extrapolation over temporal knowledge graph via hyperbolic embedding

Jia

Wang

et al. 2023

CAAI Trans on Intel Tech

View full text Add to dashboard Cite

show abstract

“…Existing datasets for event forecasting are different cropped versions of GDELT [28] and ICEWS [34]. For example, among the datasets used by current works [15,16,22,29,30], ICEWS14, ICEWS18, ICEWS05-15 include events in the ICEWS dataset of year 2014, 2018, and 2005-2015, respectively; and GDELT covers January 2018 of the original GDELT dataset. However, all of these versions only use the existing quadruple data while overlooking the context information.…”

Section: Dataset Constructionmentioning

confidence: 99%

“…With the contextspecific and cross-context modeling modules, we learn the entity and relation representations that not only capture context-aware characteristics but also preserve transferred knowledge from other contexts. Following the established approach to event forecasting [29,30], we devise a decoder based on ConvTransE [38]. In particular, given a query quadruple (𝑠, 𝑟, 𝑡, 𝑐), we first use a Con-vTransE to produce the query's representation, then score the candidate entities E via inner-product between the query and candidate representations.…”

Section: Cross-context Modelingmentioning

confidence: 99%

“…These methods consider both relational and temporal information for link prediction in the next timestamp. We consider the following SOTA methods: TANGO [16], RE-NET [22], RE-GCN [30], EvoKG [36], and HiSMatch [29]. • Temporal event forecasting methods with texts incorporate textual information into the event forecasting model, while the standard temporal KG forecasting methods only use structural information.…”

Section: Experimental Settingsmentioning

confidence: 99%

“…Model EG IR IS MRR HIT@1 HIT@3 HIT@10 MRR HIT@1 HIT@3 HIT@10 MRR HIT@1 HIT@3 HIT@10 DistMult [49] 0.1164 0.0344 0.1214 0.2927 0.1349 0.0392 0.1468 0.3379 0.1031 0.0223 0.0929 0.2950 ConvE [12] 0.1151 0.0312 0.1272 0.2882 0.1365 0.0409 0.1485 0.3400 0.1060 0.0251 0.0984 0.2935 ConvTransE [38] 0.1205 0.0377 0.1305 0.2921 0.1405 0.0462 0.1529 0.3412 0.1079 0.0287 0.0994 0.2930 RotatE [40] 0.0892 0.0125 0.0772 0.2748 0.1055 0.0125 0.1074 0.3152 0.0879 0.0132 0.0714 0.2638 RGCN [37] 0.0974 0.0279 0.1046 0.2377 0.1185 0.0366 0.1301 0.2860 0.0861 0.0242 0.0652 0.2307 TANGO [16] 0.1043 0.0240 0.1106 0.2761 0.1249 0.0281 0.1367 0.3314 0.0972 0.0171 0.0852 0.2889 RE-NET [22] 0.1212 0.0413 0.1224 0.2932 0.1401 0.0451 0.1501 0.3452 0.1064 0.0263 0.1016 0.2894 RE-GCN [30] 0.1245 0.0352 0.1366 0.3101 0.1647 0.0622 0.1796 0.3838 0.1301 0.0408 0.1281 0.3346 EvoKG [36] 0.0797 0.0012 0.0775 0.2529 0.0892 0.0011 0.0767 0.3120 0.0779 0.0008 0.0518 0.2789 HiSMatch [29] 0…”

Section: Performance Comparison (Rq1)unclassified

See 2 more Smart Citations

Context-aware Event Forecasting via Graph Disentanglement

Ma,

Ye,

et al. 2023

Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining

View full text Add to dashboard Cite

Event forecasting has been a demanding and challenging task throughout the entire human history. It plays a pivotal role in crisis alarming and disaster prevention in various aspects of the whole society. The task of event forecasting aims to model the relational and temporal patterns based on historical events and makes forecasting to what will happen in the future. Most existing studies on event forecasting formulate it as a problem of link prediction on temporal event graphs. However, such pure structured formulation suffers from two main limitations: 1) most events fall into general and high-level types in the event ontology, and therefore they tend to be coarse-grained and offers little utility which inevitably harms the forecasting accuracy; and 2) the events defined by a fixed ontology are unable to retain the out-of-ontology contextual information.To address these limitations, we propose a novel task of contextaware event forecasting which incorporates auxiliary contextual information. First, the categorical context provides supplementary fine-grained information to the coarse-grained events. Second and more importantly, the context provides additional information towards specific situation and condition, which is crucial or even determinant to what will happen next. However, it is challenging to properly integrate context into the event forecasting framework, considering the complex patterns in the multi-context scenario. Towards this end, we design a novel framework named Separation and Collaboration Graph Disentanglement (short as SeCoGD) for context-aware event forecasting. In the separation stage, we leverage the context as a prior guidance to disentangle the event graph into multiple sub-graphs, followed by a context-specific module to model the relational-temporal patterns within each context. In the collaboration stage, we design a cross-context module to retain the collaborative associations among multiple contexts. Since there is no available dataset for this novel task, we construct three largescale datasets based on GDELT. Experimental results demonstrate that our model outperforms a list of SOTA methods. The dataset and code are released via https://github.com/yecchen/SeCoGD.

show abstract