Skip to main content

Advertisement

Table 1 Scalability experiment results

From: SIA: a scalable interoperable annotation server for biomedical named entities

Parallelism Processing time (s) Throughput (doc/s) Improvement
1 8.151 25  
2 4.551 46 \(1.79{\times }\)
3 3.412 61 \(2.39{\times }\)
4 3.032 68 \(2.69{\times }\)
5 2.712 77 \(3.01{\times }\)
  1. Processing times with varying degree of parallelism, analyzing 207.551 PubMed articles with all internal annotators (SETH, mirNer, Linnaues, Banner, DiseaseNer) and ChemSpot using a single instance of SIA