Pinlong-Zhao
Pinlong-Zhao
> tf.flags.DEFINE_string('embedding_file_path', datas[use_data+"_embedding"], 'embedding file'),这句话也有问题 在你那个issue下,我回答了哈。
> 好的,谢谢,有劳了 > […](#) > ------------------ 原始邮件 ------------------ 发件人: "Pinlong-Zhao"; 发送时间: 2019年7月30日(星期二) 中午1:46 收件人: "Pinlong-Zhao/SDGCN"; 抄送: " ";"Comment"; 主题: Re: [Pinlong-Zhao/SDGCN] No targets embedding file?! (#1) tf.flags.DEFINE_string('embedding_file_path', datas[use_data+"_embedding"], 'embedding file'),这句话也有问题...
您好。这里需要注意纵坐标是aspects的数量(因为准确率也是根据aspects的情感分析对错得到的)。所以假如一个句子有两个aspects,那么一个句子中有两个aspects的aspects的个数就是2。不知道我的解释您明白了吗。
哦~这个我这里没有了。因为疫情的缘故我没回学校。但是,这代码是很简单的:用一个for循环(循环所有句子),判断句子的aspects数据,求和。就可以了。
客气啦客气啦。你的回复真快啊,哈哈~
I'm sorry to be so late reply to you. Actually, I did not do the comparison experiment. However, based on my previous experience. I think the performance of the two...
sorry,才看到。我运行的时候没有出现这个错误。你可以试着按如下方式改一下: 119行 fp = open(w2v_file,'r', encoding='UTF-8')
I have the same question with swamiviv.I read the values like garbage,too. but i can't open the link of "the sample datum converter". Can you give it again? And the...
My colleagues and I still can't open the link, do you have other ways to solve this matter, I am looking forward to your reply, thank you @kyamagu
@kyamagu Thank you for your replay. I have asked many friends to open the link.But none of them can open it. So the problem should be the network in China...