Hierarchical Representations of Freebase Topics

The dataset contains more than 21M hierarchical relationships about ≈10M topics extracted from Freebase knowledgebase. The topics span the various categories of Freebase including Science & Technology, Arts & Entertainment, Sports, Society, Products & Services, Transportation, Time & Space, Special Interests, and Commons. The relationships describe the hierarchies of topics in terms of Types, Domains, and Categories. For example, ‘Albert Einstein’ can be found as a topic that is a sub-class of ‘Person’, belonging to the ‘People’ domain and ‘Society’ category. While another entity named as ‘Albert Einstein’ can also be found as a sub-class of ‘Book’, belonging to the ‘Books’ domain and ‘Arts & Entertainment’ category. The dataset is published in JSON and CSV formats, sample files are provided to help explore how the dataset is structured. The dataset is believed to be useful for studying the inter-related connections among topics in different domains of knowledge. The first author may be contacted at ([e-mail]) for more information. The following paper may kindly be cited in case of using the dataset. Mahmoud Elbattah, Mohamed Roushdy, Mostafa Aref, Abdel-Badeeh M. Salem. “Large-Scale Entity Clustering Using Graph-Based Structural Similarity within Knowledge Graphs”, Big Data Analytics: Tools, Technology for Effective Planning, CRC Press. https://www.researchgate.net/publication/321716589_Large-Scale_Entity_Clustering_Based_on_Structural_Similarity_within_Knowledge_Graphs