Chinese NER with Height-Limited Constituent Parsing

Authors: Rui Wang, Xin Xin, Wei Chang, Kun Ming, Biao Li, Xin Fan7160-7167

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on the Onto Notes 4.0 dataset have demonstrated that the proposed model outperforms the state-of-the-art method by 2.79 points in the F1-measure.
Researcher Affiliation Collaboration Rui Wang,1 Xin Xin,1 Wei Chang,1 Kun Ming,1 Biao Li,2 Xin Fan2 1BJ ER Center of HVLIP&CC, School of Comp. Sci. & Tech., Beijing Institute of Technology, Beijing, China 2Tencent, Beijing, China {ruicn,xxin,2220160514,2120171045}@bit.edu.cn, {biotli,hsinfan}@tencent.com
Pseudocode No The paper describes algorithms (dynamic programming, pruning) but does not include a clearly labeled pseudocode or algorithm block.
Open Source Code No The paper does not provide an unambiguous statement or a link indicating that the source code for the described methodology is publicly available.
Open Datasets Yes The experiments are conducted on the dataset of Onto Notes 4.0 (Weischedel et al. 2011).
Dataset Splits Yes The dataset contains 15,724 sentences in the training set, 4,301 sentences in the development set, and 4,346 sentences in the testing set, with more than 490,000 characters in total.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments, such as GPU or CPU models.
Software Dependencies No The paper mentions using systems like ZPar and Lattice LSTM, but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes Table 1: Hyper-parameters. Some of them are employed from the work of (Zhang and Yang 2018), and others are set according to parameter analysis.