End-to-End Extraction of Structured Information from Business Documents with Pointer-Generator Networks - Archive ouverte HAL
Conference Papers Year : 2020

End-to-End Extraction of Structured Information from Business Documents with Pointer-Generator Networks

Alex Aussem
Haytham Elghazel
Jérémy Espinas
  • Function : Author
  • PersonId : 1078670

Abstract

The predominant approaches for extracting key information from documents resort to classifiers predicting the information type of each word. However, the word level ground truth used for learning is expensive to obtain since it is not naturally produced by the extraction task. In this paper, we discuss a new method for training extraction models directly from the textual value of information. The extracted information of a document is represented as a sequence of tokens in the XML language. We learn to output this representation with a pointer-generator network that alternately copies the document words carrying information and generates the XML tags delimiting the types of information. The ability of our end-to-end method to retrieve structured information is assessed on a large set of business documents. We show that it performs competitively with a standard word classifier without requiring costly word level supervision.
Fichier principal
Vignette du fichier
End_to_End_Learning_for_Extracting_Structured_Information_from_Business_Documents.pdf (1.83 Mo) Télécharger le fichier
presentation_slides.pdf (3.21 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-02958913 , version 1 (06-10-2020)

Identifiers

  • HAL Id : hal-02958913 , version 1

Cite

Clément Sage, Alex Aussem, Véronique Eglin, Haytham Elghazel, Jérémy Espinas. End-to-End Extraction of Structured Information from Business Documents with Pointer-Generator Networks. EMNLP 2020 Workshop on Structured Prediction for NLP, Nov 2020, Punta Cana (online), Dominican Republic. ⟨hal-02958913⟩
415 View
492 Download

Share

More