Search
Abstract
This paper continues the work by Wang et al. [17]. Its goal is to verify the robustness of the NGCF (Neural Graph Collaborative Filtering) technique by assessing its ability to generalize across different datasets. To achieve this, we first replicated the experiments conducted by Wang et al. [17] to ensure that their replication package is functional. We received sligthly better results for ndcg@20 and somewhat poorer results for recall@20, which may be due to the randomness. Afterward, we applied their framework to four additional datasets (NYC2014, TOKYO2014, Yelp2022, and MovieLens1M) and compared NGCF with HOP-Rec [18] and MF-BPR [14] as in the original study. Our results confirm that NGCF outperforms other models in terms of ndcg@20. However, when considering recall@20, either HOP-Rec or MF-BPR performed better on the new datasets. This finding suggests that NGCF may have been optimized for the datasets used in the original paper. Furthermore, we analyzed the models’ performance using recall@K and ndcg@K, where K was set to 1, 5, 10, and 40. The obtained results support our previous finding. The replication package for this paper can be found in our GitHub repository [1].
Citations
-
1
CrossRef
-
0
Web of Science
-
0
Scopus
Authors (5)
Cite as
Full text
full text is not available in portal
Keywords
Details
- Category:
- Monographic publication
- Type:
- rozdział, artykuł w książce - dziele zbiorowym /podręczniku w języku o zasięgu międzynarodowym
- Language:
- English
- Publication year:
- 2023
- Bibliographic description:
- Kobiela D., Groth J., Sieczczyński M., Wolniak R., Pastuszak K.: Neural Graph Collaborative Filtering: Analysis of Possibilities on Diverse Datasets// New Trends in Database and Information Systems/ : , 2023,
- DOI:
- Digital Object Identifier (open in new tab) 10.1007/978-3-031-42941-5_54
- Sources of funding:
-
- Project -
- Verified by:
- Gdańsk University of Technology
seen 13 times
Recommended for you
A new multi-process collaborative architecture for time series classification
- Z. Xiao,
- X. Xu,
- H. Zhang
- + 1 authors