Summary of Amcen: An Attention Masking-based Contrastive Event Network For Two-stage Temporal Knowledge Graph Reasoning, by Jing Yang et al.
AMCEN: An Attention Masking-based Contrastive Event Network for Two-stage Temporal Knowledge Graph Reasoning
by Jing Yang, Xiao Wang, Yutong Wang, Jiawei Wang, Fei-Yue Wang
First submitted to arxiv on: 16 May 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed attention masking-based contrastive event network (AMCEN) aims to improve the accuracy of temporal knowledge graph reasoning by addressing the imbalance between new and recurring events in datasets. The network uses a two-stage approach, first predicting future events based on local-global temporal patterns and then refining predictions using a contrastive event classifier. By incorporating historical and non-historical attention mask vectors, AMCEN alleviates the attention bias towards historical entities, allowing for more accurate reasoning. Experimental results on four benchmark datasets demonstrate the superiority of AMCEN, with notable improvements in Hits@1 indicating precise predictions about future occurrences. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary AMCEN is a new way to make more accurate predictions about what might happen next based on what happened before. Right now, there’s a problem with how we look at this data – it gets biased towards things that have already happened, which makes it harder to predict new events. AMCEN tries to fix this by looking at both historical and non-historical information, and using patterns from the past to make more informed guesses about what might happen in the future. |
Keywords
» Artificial intelligence » Attention » Knowledge graph » Mask
