TopicKA: Generating Commonsense Knowledge-Aware Dialogue Responses Towards the Recommended Topic Fact

Authors: Sixing Wu, Ying Li, Dawei Zhang, Yang Zhou, Zhonghai Wu

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on an open Chinese benchmark dataset indicate that our model outperforms baselines in terms of both the objective and the subjective metrics. We collected and constructed a large-scale Chinese commonsense knowledge graph. Experimental results indicate our Topic KA outperforms various kinds of baselines.
Researcher Affiliation Academia 1School of Electronics Engineering and Computer Science, Peking University, Beijing, China 2National Research Center of Software Engineering, Peking University, Beijing, China 3Auburn University, Auburn, Alabama, USA
Pseudocode No The paper describes the model architecture and mathematical formulations but does not include any explicitly labeled pseudocode blocks or algorithms.
Open Source Code Yes Our experimental resources are open released1. 1https://github.com/pku-orangecat/IJCAI2020-Topic KA
Open Datasets Yes Our approach is evaluated on an open Chinese benchmark dataset [Li and Yan, 2018], which is collected from the largest Chinese SNS (weibo.com). We collect the commonsense knowledge from the Concept Net (conceptnet.io).
Dataset Splits Yes After the alignment, the remaining aligned data are randomly divided into three sets: training set, validation set, and test set, which have 847K, 30K, and 30K pairs of dialogue, respectively.
Hardware Specification No The paper does not provide specific hardware details such as GPU/CPU models, memory, or cloud instance types used for the experiments. It only mentions software and hyperparameters like "batch size is 50, word embedding dimension is 300, GRU dimension is 512, Adam with the initial learning rate 0.0001 is used to optimize, and the maximum epoch is limited to 20."
Software Dependencies No The paper mentions that "other models are re-implemented by Tensorflow." but does not provide specific version numbers for TensorFlow or any other software dependencies.
Experiment Setup Yes Most hyper parameters are selected from the CCM: batch size is 50, word embedding dimension is 300, GRU dimension is 512, Adam with the initial learning rate 0.0001 is used to optimize, and the maximum epoch is limited to 20.