search

yingtaoluo / Spatial-Temporal-Attention-Network-for-POI-Recommendation

Codes for a WWW'21 Paper. POI recommender system for location/trajectory prediction.
https://doi.org/10.1145/3442381.3449998
164 stars 39 forks source link

Call for the NYC dataset #30

Closed swiftCC closed 2 months ago

swiftCC commented 5 months ago

hello, the link below can't be open. Can u send me the origin NYC dataset that you have used in the experiment? thanks!

yingtaoluo commented 5 months ago

Thanks. Please see the data folder of this repo.

Iscodebrid commented 4 months ago

Thanks. Please see the data folder of this repo.

Hello, may I ask if you can share the other two datasets (.pkl format) besides the NYC dataset? Thanks

yingtaoluo commented 4 months ago

I directly re-use the data used by the STRNN paper in 2016, and I also do not have the code to get the csv file. The npy file is the same to the csv file. The pkl file is too large and I suggest using the load.py to directly generate from the original data.

Iscodebrid commented 4 months ago

I directly re-use the data used by the STRNN paper in 2016, and I also do not have the code to get the csv file. The npy file is the same to the csv file. The pkl file is too large and I suggest using the load.py to directly generate from the original data.

Thank you for your reply! I would like to ask you if you have any suggestions on applying this model to a federated learning framework?

yingtaoluo commented 4 months ago

I directly re-use the data used by the STRNN paper in 2016, and I also do not have the code to get the csv file. The npy file is the same to the csv file. The pkl file is too large and I suggest using the load.py to directly generate from the original data.

Thank you for your reply! I would like to ask you if you have any suggestions on applying this model to a federated learning framework?

Thank you for the feedback. I am not an expert in federated learning, but many models can be extended to a federated learning version. I could only guess that if there is a very suitable problem, such as if different institutions have different geo data and want to protect their own data privacy, it will make a lot of sense to do so.

Iscodebrid commented 4 months ago

Thank you for your reply, it will help me a lot！