IMN
1.0.0
该存储库包含 Gu 等人发表的 CIKM 2019 论文《Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots》的源代码和数据集。
我们提出的交互式匹配网络(IMN)在四个可公开用于多轮对话研究的大型数据集上取得了最先进的性能。
Python 2.7
张量流1.4.0
您可以在此处下载我们论文中使用的处理后的数据集并将其解压缩到data
文件夹中。
Ubuntu_V1
Ubuntu_V2
豆瓣
电子商务
以Ubuntu_V1为例。
cd scripts
bash ubuntu_train.sh
训练过程记录在log_train_IMN_UbuntuV1.txt
文件中。
bash ubuntu_test.sh
测试过程记录在log_test_IMN_UbuntuV1.txt
文件中。您可以获得一个ubuntu_test_out.txt
文件,该文件记录每个上下文响应对的分数。运行以下命令,您可以计算 Recall 的指标。
python compute_recall.py
如果您使用代码和数据集,请引用以下论文: “Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots” Jia-Chen Gu、Zhen-Hua Ling、Quan Liu。 CIKM (2019)
@inproceedings{Gu:2019:IMN:3357384.3358140,
author = {Gu, Jia-Chen and
Ling, Zhen-Hua and
Liu, Quan},
title = {Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots},
booktitle = {Proceedings of the 28th ACM International Conference on Information and Knowledge Management},
series = {CIKM '19},
year = {2019},
isbn = {978-1-4503-6976-3},
location = {Beijing, China},
pages = {2321--2324},
url = {http://doi.acm.org/10.1145/3357384.3358140},
doi = {10.1145/3357384.3358140},
acmid = {3358140},
publisher = {ACM},
}