Skip to main content

Vocabulary of natural language processing

Search from vocabulary

Concept information

Preferred term

BANNER  

Definition

  • A system for the recognition of named entities intended primarily for biomedical texts. It uses conditional random fields as the main recognition engine and includes a wide range of the best techniques. described in recent literature. (Loterre)

Broader concept

Definitional context(s)

  • BANNER is a freely available tool for identifying gene mentions. (Kim, Kim, Comeau & Wilbur, 2012)

Example

  • A BANNER input file contains a text which has been separated into sentences. (Khordad, Mercer & Rogan, 2012)
  • BANNER is an open-source biomedical NER system implemented using conditional random fields (CRFs) (Lafferty et al. (Tang, Zhang, Zhang, Wu & Zhuang, 2017)
  • Even though BANNER covers most popular feature types it does not apply semantic features or other post-processing procedures such as abbreviation processing. (Kim, Kim, Comeau & Wilbur, 2012)
  • The tokenization process in BANNER breaks tokens into either a contiguous block of letters and/or digits or a single punctuation mark. (Khordad, Mercer & Rogan, 2012)

In other languages

URI

http://data.loterre.fr/ark:/67375/8LP-D2V4QCK4-K

Download this concept:

RDF/XML TURTLE JSON-LD Last modified 5/3/24