Question for result

[Ricca-xie]

Sorry for disturb, I have run your code by using the default parameters, but obtain 2-14 percent lower f1 score than that recorded in your paper result. Another problem is that I couldn't run the code for Walmart-Amazon3. Hope you could give me some suggestions.

[cipnlu]

Sorry for disturb, I have run your code by using the default parameters, but obtain 2-14 percent lower f1 score than that recorded in your paper result. Another problem is that I couldn't run the code for Walmart-Amazon3. Hope you could give me some suggestions.

For the first question, you can try different manual seeds in your experiments. For the second question, we have fixed the file names of this dataset, therefore it can work now.

[Ricca-xie]

Thanks so much.

[LiuMiaoo]

Sorry for disturb, I have run your code by using the default parameters, but obtain 2-14 percent lower f1 score than that recorded in your paper result. Another problem is that I couldn't run the code for Walmart-Amazon3. Hope you could give me some suggestions.

Modify the data set to data/walmart_amazon_3, and an error assert num_left == num_right will be reported. How do you solve this problem?

[LiuMiaoo]

Sorry for disturb, I have run your code by using the default parameters, but obtain 2-14 percent lower f1 score than that recorded in your paper result. Another problem is that I couldn't run the code for Walmart-Amazon3. Hope you could give me some suggestions.

For the first question, you can try different manual seeds in your experiments. For the second question, we have fixed the file names of this dataset, therefore it can work now.

Changing the data set to data/walmart_amazon_3 will report an error assert num_left == num_right. How should I solve this problem?

[LearnerHong]

抱歉打扰，我已经使用默认参数运行了您的代码，但获得的 f1 分数比您论文结果中记录的低 2-14%。另一个问题是我无法运行 Walmart-Amazon3 的代码。希望你能给我一些建议。

哈哈哈，低这么多吗，，