Baseline

[CVxTz]

Hello ! 

Is there any baseline code available ? 
The training data format is a bit confusing and it would be really helpful to have a simple baseline code to get started with.

Best,

CVxTz 

[kochki...@gmail.com]

Hi,

I am working on it at the moment. 

In the mean time you might find the code in https://github.com/kochkinaelena/branchLSTM helpful for processing tweets. 

You can have a look at both master (uses lasagna) and keras branches (keras branch is not documented yet). 

Best regards,

Elena

вторник, 25 сентября 2018 г., 15:00:07 UTC+1 пользователь CVxTz написал:

[CVxTz]

Thanks  Elena ! 

[kochki...@gmail.com]

Hi, 

I have just shared my baseline code for both tasks on github. https://github.com/kochkinaelena/RumourEval2019

Hope it helps. Let me know if you have any questions.

Best,

Elena

вторник, 25 сентября 2018 г., 15:44:35 UTC+1 пользователь CVxTz написал:

Thanks  Elena ! 

[Petr Ostroukhov]

Hi, is there any paper about the method, used in baseline for subtask B? In RumourEval 2019: Determining Rumour Veracity and Support for Rumours it's said, that for this subtask the SOTA baseline implementation from Enayet and El-Beltagy (RumourEval 2017) would be provided. However they didn't use any neural nets, there are LSTMs in your implementation.

пятница, 19 октября 2018 г., 18:28:28 UTC+3 пользователь kochki...@gmail.com написал:

[kochki...@gmail.com]

Hi,

Sorry for such late response. I think there have been some mis-communication and I thought that any baseline would do. 

My implementation of Enayet and El-Beltagy (RumourEval 2017) have now been added to the github repo. 

Best,

Elena

понедельник, 10 декабря 2018 г., 16:11:38 UTC пользователь Petr Ostroukhov написал: