search

tsinghua-fib-lab / CLSR

The official implementation of "Disentangling Long and Short-Term Interests for Recommendation" (WWW '22)
MIT License
70 stars 19 forks source link

comparison experiments #13

Closed lypla closed 2 years ago

lypla commented 2 years ago

Hello, the kuaishou dataset has not been published yet. Apart from Taobao dataset,are there any other datasets available for comparison experiments in this code? Any suggestions?Thanks!

DavyMorgan commented 2 years ago

Hello, you can use the provided data preprocess codes to generate training and test data for your custom datasets. The data preprocess codes can be found at here. Most of the datasets adopted in sequential recommendation can be used, such as Amazon, Yelp, etc. Good Luck!

lypla commented 2 years ago

Thank you very much. When processing the yelp dataset, I noticed that in the sequential_review.py, load_data in the yelp_main uses categories.json, as follows: “ with open(os.path.join(dirs, 'categories.json'), 'r') as f: category = json.load(f) ” When I downloaded the yelp dataset, I found no categories.json.Is the yelp dataset not completely downloaded? (My download address:https://www.yelp.com/dataset/download）

DavyMorgan commented 2 years ago

You can find the category file of the Yelp dataset at here.