TachiChan / IJCAI2019_HGAT

Semi-supervised User Profiling with Heterogeneous Graph Attention Networks, IJCAI 2019.
47 stars 11 forks source link

Input features for baselines #4

Open FeiWang96 opened 4 years ago

FeiWang96 commented 4 years ago

Interesting task!

However, the results of baselines are confusing. SVM, LR, GCN and GAT only achieve accuracies around 0.5 for the binary classification (gender classification) task in your paper.

Could you please show the detail of baselines?

hetianqi210701 commented 4 years ago

他给的数据集不是user,user_click这几种格式吗,怎么转为npy格式的,是我错过了哪里吗

FeiWang96 commented 4 years ago

他给的数据集不是user,user_click这几种格式吗,怎么转为npy格式的,是我错过了哪里吗

论文里有详细的处理说明,可以据此复现。

hetianqi210701 commented 4 years ago

好的,谢谢,我在好好看看,非常感谢

------------------ 原始邮件 ------------------ 发件人: "Fei Wang"<notifications@github.com>; 发送时间: 2020年6月3日(星期三) 中午12:37 收件人: "TachiChan/IJCAI2019_HGAT"<IJCAI2019_HGAT@noreply.github.com>; 抄送: "普通"<2371750994@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [TachiChan/IJCAI2019_HGAT] Input features for baselines (#4)

他给的数据集不是user,user_click这几种格式吗,怎么转为npy格式的,是我错过了哪里吗

论文里有详细的处理说明,可以据此复现。

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

gu18168 commented 4 years ago

@FeiWang96 你好,能提供一下你的复现代码参考一下吗,麻烦了

XIAOHUOCHAI123 commented 3 years ago

你好,可以提供数据处理的方法吗,谢谢!