mjugo
commented
4 years ago Hi Yifei,
sorry for the late reply. It seems I seems the notification mail I got from github went the the spam folder.
I don't know if this info is of any use to you still, but when doing the analyses for the paper, we employed the GRU4Rec strategy from another recommender library, to which we connected via a webservice interface.
The author of this library can probably better answer your question since this functionality was implemented in their paper. The source code for this library can be found here: https://github.com/rn5l/session-rec
Best, Michael
yifeim
Hi Michael,
We were reading your paper and hoping to find the GRU4Rec evaluation, particularly: "For GRU4Rec, we employed a heuristic to sample from the more recent sessions to take recent temporal shifts into account, as proposed in [45]."
In [45], We were able to find the following:
"We propose a simple solution to get the best of both worlds via pre-training. We first train a model on the entire dataset. The trained model is then used to initialize a new model, which is only trained using only a more recent subset of the data, e.g. the last month worth of data out of a year of click sequences. This allows the model to have the benefit of a good initialization using large amounts of data, and yet is focused on more recent click-sequences. In this way, it resembles the fine-tuning process used in training of image-based networks [2], where the models are typically initialized by pre-training on ImageNet (a large image classification dataset) before the weights are fine-tuned on a smaller image dataset in the desired domain."
How was the sampling done in particular? Are there open-source template codes?
Thanks.