RaDE+: A semantic rank-based graph embedding algorithm
Due to the possibility of capturing complex relationships existing between nodes, many applications benefit from being modeled with graphs. However, performance issues can be observed in large-scale networks, making it computationally unfeasible to process in various scenarios. Graph Embedding metho...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2022 |
| País: | Brasil |
| Institución: | Universidade Estadual Paulista (UNESP) |
| Repositorio: | Repositório Institucional da UNESP |
| Idioma: | inglés |
| OAI Identifier: | oai:repositorio.unesp.br:11449/241880 |
| Acceso en línea: | http://dx.doi.org/10.1016/j.jjimei.2022.100078 http://hdl.handle.net/11449/241880 |
| Access Level: | acceso abierto |
| Palabra clave: | Diffusion Graph embedding Interpretability Network representation learning Ranking Semantic Unsupervised |
| Sumario: | Due to the possibility of capturing complex relationships existing between nodes, many applications benefit from being modeled with graphs. However, performance issues can be observed in large-scale networks, making it computationally unfeasible to process in various scenarios. Graph Embedding methods emerge as a promising solution for finding low-dimensional vector representations for graphs, preserving their original properties such as topological characteristics, affinity, and shared neighborhood between nodes. Based on the vectorial representations, retrieval and machine learning techniques can be exploited to execute tasks such as classification, clustering, and link prediction. In this work, we propose RaDE (Rank Diffusion Embedding), an effective and efficient approach that considers rank-based graphs and representative nodes selection for learning a low-dimensional vector. We also present RaDE+, a variant that considers multiple representative nodes for more robust representations. The proposed approach was evaluated on 8 network datasets, including social, co-reference, textual, and image networks, with different properties. Vector representations generated with RaDE achieved effective results in visualization and retrieval tasks when compared to vector representations generated by other recent related methods. |
|---|