Neural Snowball for Few-Shot Relation Learning

Authors: Tianyu Gao, Xu Han, Ruobing Xie, Zhiyuan Liu, Fen Lin, Leyu Lin, Maosong Sun7772-7779

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments show that our model can gather high-quality instances for better fewshot relation learning and achieves significant improvement compared to baselines.
Researcher Affiliation Collaboration 1Department of Computer Science and Technology, Tsinghua University, Beijing, China Institute for Artificial Intelligence, Tsinghua University, Beijing, China Beijing National Research Center for Information Science and Technology 2Search Product Center, We Chat Search Application Department, Tencent, China
Pseudocode Yes Algorithm 1 describes the process. The fine-tuning process is used as one of our baselines. We also adopt this algorithm in each step of Neural Snowball after gathering new instances in Sr. [...] Algorithm 1: Fine-tuning the Classifier
Open Source Code Yes Codes and datasets are released on https://github.com/thunlp/Neural-Snowball.
Open Datasets Yes For now the only qualified dataset is Few Rel (Han et al. 2018). It contains 100 relations and 70,000 instances from Wikipedia.
Dataset Splits Yes The dataset is divided into three subsets: training set (64 relations), validation set (16 relations) and test set (20 relations).
Hardware Specification No The paper does not specify the exact hardware used for experiments, such as specific CPU/GPU models or memory.
Software Dependencies No The paper mentions 'Adam' as an optimizer and 'CNN' and 'BERT' as encoders but does not provide specific version numbers for these or any other software dependencies.
Experiment Setup Yes For the fine-tuning, after grid searching, we adopt training epochs e = 50, batch size bs = 10, learning rate λ = 0.05 and negative loss coefficient μ = 0.2. BERT fine-tuning shares the same parameters except for λ = 0.01 and μ = 0.5. For the Neural Snowball process, we also determine our parameters by grid searching. We set K1 and K2, the numbers of added instances for each stage, as 5, and the thresholds of RSN for each stage, α and β, as 0.5. We adopt 0.9 for the classifier threshold θ.