Table-to-Text Generation by Structure-Aware Seq2seq Learning

Authors: Tianyu Liu, Kexiang Wang, Lei Sha, Baobao Chang, Zhifang Sui

AAAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct experiments on the WIKIBIO dataset which contains over 700k biographies and corresponding infoboxes from Wikipedia. ... Automatic evaluations also show our model outperforms the baselines by a great margin.
Researcher Affiliation Academia Tianyu Liu, Kexiang Wang, Lei Sha, Baobao Chang, Zhifang Sui Key Laboratory of Computational Linguistics, Ministry of Education, School of Electronics Engineering and Computer Science, Peking University, Beijing, China {tianyu0421, wkx, shalei, chbb, szf}@pku.edu.cn
Pseudocode No The paper includes mathematical equations and architectural diagrams but no structured pseudocode or algorithm blocks labeled as such.
Open Source Code Yes Code for this work is available on https://github.com/tyliupku/wiki2bio.
Open Datasets Yes We use WIKBIO dataset proposed by Lebret, Grangier, and Auli (2016) as the benchmark dataset. WIKBIO contains 728,321 articles from English Wikipedia (Sep 2015).
Dataset Splits Yes The corpus has been divided in to training (80%), testing (10%) and validation (10%) sets.
Hardware Specification No The paper discusses model parameters and experimental setup (e.g., word dimension, hidden size, batch size, optimizer) but does not provide specific hardware details like GPU/CPU models or other computing infrastructure used for experiments.
Software Dependencies Yes We use the Ken LM toolkit to train 5-gram models without pruning.
Experiment Setup Yes The detail of model parameters is listed in Table 2. Word dimension 400 Field dimension 50 Position dimension 5 Hidden size 500 Batch size 32 Learning rate 0.0005 Optimizer Adam