Query-Based Keyphrase Extraction from Long Documents

Autores/as

  • Martin Dočekal Brno University of Technology
  • Pavel Smrž

DOI:

https://doi.org/10.32473/flairs.v35i.130737

Palabras clave:

keyphrase, keyword, long documents, query-based keyphrase extraction, BERT, transformer

Resumen

Transformer-based architectures in natural language processing force input size limits that can be problematic when long documents need to be processed. This paper overcomes this issue for keyphrase extraction by chunking the long documents while keeping a global context as a query defining the topic for which relevant keyphrases should be extracted. The developed system employs a pre-trained BERT model and adapts it to estimate the probability that a given text span forms a keyphrase. We experimented using various context sizes on two popular datasets, Inspec and SemEval, and a large novel dataset. The presented results show that a shorter context with a query overcomes a longer one without the query on long documents.

Descargas

Publicado

2022-05-04

Cómo citar

Dočekal, M., & Smrž, P. (2022). Query-Based Keyphrase Extraction from Long Documents. The International FLAIRS Conference Proceedings, 35. https://doi.org/10.32473/flairs.v35i.130737

Número

Sección

Special Track: Applied Natural Language Processing