search

WangYuhan-0520 / Amazon-KG-v2.0-dataset

The Amazon-KG dataset is a KG-enhanced cross-domain recommendation dataset, which is constructed based on the widely used dataset [Amazon](https://nijianmo.github.io/amazon/index.html) and the well-known KG [DBpedia](http://dbpedia.org), aiming to introduce KG for cross-domain linking and knowledge enhancement.
4 stars 0 forks source link

Questions about datasets #1

Open zhangyuge666 opened 2 weeks ago

zhangyuge666 commented 2 weeks ago

Hello, I would like to ask whether the source of the movie data set is Amazon's data set in 2018 or 2023. Because I downloaded the data set in 2018, I can't find the corresponding item in the original data set by using the item name in your data, and user can only find part of it.

WangYuhan-0520 commented 2 weeks ago

Hello! We used Amazon Review Data (2018) from https://nijianmo.github.io/amazon/index.html and linked the dataset's items to the DBpedia Knowledge Graph. It is important to note that some items cannot find corresponding KG entities in DBpedia, and we only keep the items with aligned entities and their related interactions in this dataset, so the dataset is inevitably downsized. Thank you for your attention!

zhangyuge666 commented 1 week ago

Thank you for your reply. I have solved the problem. I want to know whether your paper has been pre-printed or officially published, because my research work uses a data set published by you, and I want to quote your paper in the introduction part of my paper data set.

WangYuhan-0520 commented 1 week ago

Yes, this work has been accepted for publication by SIGIR2024. The title of the paper is "Amazon-KG: A Knowledge Graph Enhanced Cross-Domain Recommendation Dataset" and you can get information about it by visiting this page (https://sigir-2024. github.io/program_papers.html#rr). Thanks for the support!