Continuous Query Decomposition

---

Final Project

The final version of the project was moved to the official UCL repo. For any kind of discussions, issues and reproductions please head there.

Update

We implemented CQD in the KGReasoning framework, a library from SNAP implementing several Complex Query Answering models, which also supports experimenting with the Query2Box and BetaE datasets (in this repo, we only consider the former). Our implementation is available at this link.

---

This repository contains the initial implementation for our ICLR 2021 (Oral, Outstanding Paper Award) paper, Complex Query Answering with Neural Link Predictors:

@inproceedings{
    arakelyan2021complex,
    title={Complex Query Answering with Neural Link Predictors},
    author={Erik Arakelyan and Daniel Daza and Pasquale Minervini and Michael Cochez},
    booktitle={International Conference on Learning Representations},
    year={2021},
    url={https://openreview.net/forum?id=Mos9F9kDwkz}
}

In this work we present CQD, a method that reuses a pretrained link predictor to answer complex queries, by scoring atom predicates independently and aggregating the scores via t-norms and t-conorms.