DBLP4k

class dhg.data.DBLP4k(data_root=None)[source]

Bases: dhg.data.base.BaseData

The DBLP-4k dataset is a citation network dataset for node classification task. The dataset is an academic network from four research areas. There are 14,475 authors, 14,376 papers, and 20 conferences, among which 4,057 authors, 20 conferences and 100 papers are labeled with one of the four research areas (database, data mining, machine learning, and information retrieval). The vertice denotes author, and three types of correlation (co-paper, co-term, co-conference) can be used for building hyperedges. More details see the PathSim: Meta Path-Based Top-K Similarity Search in Heterogeneous Information Networks paper.

The content of the DBLP-4k dataset includes the following:

  • num_classes: The number of classes: \(4\).

  • num_vertices: The number of vertices: \(4,057\).

  • num_paper_edges: The number of hyperedges constructed by the co-paper correlation: \(14,328\).

  • num_term_edges: The number of hyperedges constructed by the co-term correlation: \(7,723\).

  • num_conf_edges: The number of hyperedges constructed by the co-conference correlation: \(20\).

  • dim_features: The dimension of author features: \(334\).

  • features: The author feature matrix. torch.Tensor with size \((4,057 \times 334)\).

  • labels: The label list. torch.LongTensor with size \((4,057, )\).

  • edge_by_paper: The hyperedge list constructed by the co-paper correlation. List with length \((14,328)\).

  • edge_by_term: The hyperedge list constructed by the co-term correlation. List with length \((7,723)\).

  • edge_by_conf: The hyperedge list constructed by the co-conference correlation. List with length \((20)\).

  • paper_author_dict: The dictionary of {paper_id: [author_id, ...]}. Dict with length \((14,328)\).

  • term_paper_dict: The dictionary of {term_id: [paper_id, ...]}. Dict with length \((7,723)\).

  • conf_paper_dict: The dictionary of {conf_id: [paper_id, ...]}. Dict with length \((20)\).

Parameters

data_root (str, optional) – The data_root has stored the data. If set to None, this function will auto-download from server and save into the default direction ~/.dhg/datasets/. Defaults to None.