afshinrahimi / geographconv Goto Github PK

Hi,

Thank you for sharing the code of this impressive work.
I have two questions regarding how to reproduce the results in the paper.

1. Based on the paper, it seems that the TWITTER-WORLD dataset is larger than the TWITTER-US; however, when I downloaded the data from this link, I found that the files in the na folder is larger than the ones in the world folder, which confuses me. I wonder if there is a naming typo here.

2. I tried the following command to get the GCN results with the default hyper-paramters on GEOTEXT:
   python gcnmain.py -save -i cmu -d data/cmu -enc latin1
   Unfortunately, I only get:
   PM dev results:
   PM Mean: 565 Median: 103 Acc@161: 54
   PM test results:
   PM Mean: 578 Median: 99 Acc@161: 53
   This is a lot worse than the Mean: 546 Median: 45 Acc@161: 60 reported in the paper.
   Could you please share the commands you used to produce the amazing GCN results on all three datasets in Table 1 of the paper?

Your data uses these identifiers for users: USER_ee551c6c.

These don't correspond to the type of id's you get from twitter. How do you convert the twitter id's in this format. I am asking this because I need to augment my data to your dataset and would use a similar conversion for my data as well.

I wish to get label information out. Which label corresponds to what city?

Hi,
First of all, very impressive work that set the foundation of user geolocation task using graph-based method.

However, there is one question regarding twitter-world dataset. As the mention graph is critical in the methodology, current twitter-world dataset provided by your link seems does not contain the mention@ or reply information towards your encoded id.

For example, the user id in the dataset seems to be encoded as '421527640' , '86941247' but the mentioned user in the stacked tweet seems to be '171966527 36.09986 -80.24422 ||| @wakefan2321 what's your PSN name?', which is the plaintext. Therefore, this makes it unavailable to construct the mention graph based on the encoded user id.

Can you kindly clarify this, maybe I miss some details in the dataset.

Best,

I run the code and it came out some import errors inside the Lasagne package code. I have to edit the code of Lasagne to run the code.

==========================
I solved it by updating the Lasagne.

Hello,

I am interested in your work about the Twitter graph dataset and i'd like to do the same thing on my own side.

I need to know beforehand, how long did it took to run the whole process and on what kind of hardware ? I am not looking for performance, I just want an overall estimation about the needed time to compute in order to organize my upcoming work (and also it would help me detect a problem if it runs for too long).

Also the Twitter graph dataset is quite big so I'm being cautious, how much RAM will I need to make this run ?

Thank you for putting this into github, I mean it.

Best regards