Complex Query Answering with Neural Link Predictors
Authors: Erik Arakelyan, Daniel Daza, Pasquale Minervini, Michael Cochez
ICLR 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | In this section we demonstrate the effectiveness of CQD on the task of answering complex queries that cannot be answered using the incomplete KG, and report experimental results for continuous optimisation (CQD-CO, Section 3.1) and beam search (CQD-Beam, Section 3.2). |
| Researcher Affiliation | Collaboration | 1UCL Centre for Artificial Intelligence, University College London, United Kingdom 2Vrije Universiteit Amsterdam, The Netherlands 3University of Amsterdam, The Netherlands 4Discovery Lab, Elsevier, The Netherlands |
| Pseudocode | No | The paper describes its methods in prose (e.g., in Section 3.2 'COMPLEX QUERY ANSWERING VIA COMBINATORIAL OPTIMISATION'), but does not include any formally structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | All our source code and datasets are available online 1. 1At https://github.com/uclnlp/cqd |
| Open Datasets | Yes | Following Ren et al. (2020), we evaluate our approach on FB15k (Bordes et al., 2013) and FB15k-237 (Toutanova & Chen, 2015) two subset of the Freebase knowledge graph and NELL995 (Xiong et al., 2017), a KG generated by the NELL system (Mitchell et al., 2015). |
| Dataset Splits | Yes | Table 1: Number of queries in the datasets used for evaluation of query answering performance. [...] FB15k 273,710 [...] 59,097 [...] 67,016 |
| Hardware Specification | No | The Acknowledgements section mentions 'Finally, we thank NVIDIA for GPU donations.' However, it does not specify any particular GPU model, CPU model, or other detailed hardware specifications used for the experiments. |
| Software Dependencies | No | The paper mentions software components and algorithms used (e.g., 'Compl Ex', 'Adagrad optimiser', 'Adam'), but it does not provide specific version numbers for any of these software dependencies or programming languages (e.g., Python, PyTorch/TensorFlow versions). |
| Experiment Setup | Yes | We fix a learning rate of 0.1 and use the Adagrad optimiser. We consider ranks (size of the embedding) in {100, 200, 500, 1000}, batch size in {100, 500, 1000}, and regularisation coefficients in the interval [10^-4, 0.5]. For CQD-CO, we optimise variable and target embeddings with Adam, using the same initialisation scheme as Lacroix et al. (2018), with an initial learning rate of 0.1 and a maximum of 1,000 iterations. For CQD-Beam, the beam size k ∈ {2^2, 2^3, ..., 2^8} is found on an held-out validation set. |