ESBM: An Entity Summarization BenchMark

摘要: Entity summarization is the problem of computing an optimal compact summary for entity by selecting a size-constrained subset triples from RDF data. supports multiplicity applications and has led to fruitful research. However, there lack evaluation efforts that cover broad spectrum existing systems. One reason benchmarks evaluation. Some are no longer available, while others small have limitations. In this paper, we create Summarization BenchMark (ESBM) which overcomes limitations meets standard desiderata benchmark. Using largest available benchmark evaluating general-purpose summarizers, perform most extensive experiment date where 9 systems compared. Considering all these unsupervised, also implement evaluate supervised learning based system reference.