Collection of datasets (Karate, Football, Dolphin, PolBook, PolBlog, Email, CiteSeer, CORA, PubMed, and DBLP) for a parallel stacked autoencoder-based method driven by semi-supervised learning for community detection in social networks

Zachary Karate Club (Karate): This dataset, collected by Wayne Zachary in 1977, represents the social network of members in a university Karate club. Each node represents a club member, and edges denote contact or interaction between members. It contains 2 communities. Bottlenose Dolphin (Dolphin): This dataset represents the social interactions of 62 dolphins in New Zealand's Doubtful Sound. It is an undirected network with two main communities, where the larger group is further subdivided into smaller groups. It contains 2 communities. American Football (Football): The American Football dataset captures games played between 115 Division IA teams during the 2000 Fall season. Each node represents a team, and edges represent games between teams. The teams are grouped into 12 conferences, with more frequent games occurring within each conference. It contains 12 communities. Political Books (Polbooks): The Polbooks dataset includes 105 nodes representing books about U.S. politics sold on Amazon around the 2004 presidential election. Edges indicate that the same customers frequently bought these books together. It contains 3 communities. Political Blogs (Polblogs): This dataset includes 1,490 blogs about US politics, created by Adamic and Glance in 2005. Each node represents a blog classified as either liberal or conservative. Edges represent links between blogs. It contains 2 communities. Email-Eu (Email): The Email-Eu dataset represents email communication trends within a large European research institution. It records email exchanges between individuals over time, providing insights into the dynamics and organization of the communication network. It contains 42 communities. CiteSeer: The CiteSeer dataset is a citation network of scientific publications in the field of computer science. In this graph, each node represents a research paper, and edges denote citation relationships between papers. The documents are categorized into several predefined research topics, which are commonly used as ground-truth communities for evaluating community detection algorithms. It contains 6 communities. CORA: The CORA dataset consists of a citation network of scientific publications in the domain of machine learning. Nodes correspond to individual papers, while edges represent citation links between them. Each paper is assigned to one of several subject categories, which serve as ground-truth communities. It contains 7 communities. PubMed: The PubMed dataset is a citation network of biomedical research papers, particularly focusing on publications related to diabetes. In this network, nodes represent scientific articles and edges correspond to citation relationships. The papers are classified into three major categories according to the type of diabetes studied, which are often treated as ground-truth communities for evaluation purposes. It contains 3 communities. DBLP: The DBLP dataset is a network of academic article citations, containing information about authors, titles, and publication venues. It is commonly used for tasks such as topic modelling, influence analysis, and author clustering. It contains 13,477 communities. Network Data Statistics Karate Nodes 34 Edges 78 Density 0.139037 Maximum degree 17 Minimum degree 1 Average degree 4 Assortativity -0.475613 Number of triangles 135 Average number of triangles 3 Maximum number of triangles 18 Average clustering coefficient 0.570638 Fraction of closed triangles 0.255682 Maximum k-core 5 Network Data Statistics Dolphin Nodes 62 Edges 159 Density 0.0840825 Maximum degree 12 Minimum degree 1 Average degree 5 Assortativity -0.043594 Number of triangles 285 Average number of triangles 4 Maximum number of triangles 17 Average clustering coefficient 0.258958 Fraction of closed triangles 0.308776 Maximum k-core 5 Network Data Statistics Football Nodes 115 Edges 613 Density 0.198319 Maximum degree 19 Minimum degree 1 Average degree 6 Assortativity -0.176253 Number of triangles 351 Average number of triangles 10 Maximum number of triangles 53 Average clustering coefficient 0.338986 Fraction of closed triangles 0.329268 Maximum k-core 7 Lower bound of Maximum Clique 6 Network Data Statistics Polbook Nodes 105 Edges 441 Density 0.0807692 Maximum degree 25 Minimum degree 2 Average degree 8 Assortativity -0.127896 Number of triangles 1.7K Average number of triangles 16 Maximum number of triangles 76 Average clustering coefficient 0.487527 Fraction of closed triangles 0.348403 Maximum k-core 7 Network Data Statistics Polblog Nodes 1.5K Edges 19K Density 0.0171477 Maximum degree 467 Minimum degree 0 Average degree 25 Assortativity -0.196103 Number of triangles 459.4K Average number of triangles 308 Maximum number of triangles 9.5K Average clustering coefficient 0.312823 Fraction of closed triangles 0.251111 Maximum k-core 44 Lower bound of Maximum Clique 19 Network Data Statistics E-Mail Nodes 1k Edges 25.6k Density 0.00027 Maximum degree 1,383 Minimum degree 1 Average degree 10.02 Assortativity -0.110 Number of triangles 727,044 Average number of triangles 19.82 Maximum number of triangles 5,234 Average clustering coefficient 0.497 Fraction of closed triangles 0.085 Maximum k-core 43 Lower bound of Maximum Clique 20 Network Data Statistics CiteSeer Nodes 3.3K Edges 4.5K Density 0.000851796 Maximum degree 99 Minimum degree 1 Average degree 2 Assortativity 0.0480638 Number of triangles 3.5K Average number of triangles 1 Maximum number of triangles 85 Average clustering coefficient 0.144651 Fraction of closed triangles 0.130144 Maximum k-core 8 Lower bound of Maximum Clique 6 Network Data Statistics CORA Nodes 2.7K Edges 5.4K Density 0.0014812 Maximum degree 169 Minimum degree 1 Average degree 4 Assortativity -0.0656162 Number of triangles 5.4K Average number of triangles 2 Maximum number of triangles 170 Average clustering coefficient 0.246175 Fraction of closed triangles 0.0997136 Maximum k-core 6 Lower bound of Maximum Clique 5 Network Data Statistics PubMed Nodes 19k Edges 44k Density 0.00023 Maximum degree 171 Minimum degree 1 Average degree 4.50 Assortativity 0.145 Number of triangles 13144 Average number of triangles 0.67 Maximum number of triangles 322 Average clustering coefficient 0.633 Fraction of closed triangles 0.051 Maximum k-core 24 Lower bound of Maximum Clique 8 Network Data Statistics DBLP Nodes 317K Edges 1M Density 0.000627401 Maximum degree 710 Minimum degree 1 Average degree 7 Assortativity -0.0455856 Number of triangles 133.9K Average number of triangles 10 Maximum number of triangles 2.5K Average clustering coefficient 0.119232 Fraction of closed triangles 0.0628303 Maximum k-core 13 Lower bound of Maximum Clique 8