JSAI2024

Presentation information

Poster Session

Poster session » Poster session

[4Xin2] Poster session 2

Fri. May 31, 2024 12:00 PM - 1:40 PM Room X (Event hall 1)

[4Xin2-49] Extraction of organization names from text data including table structure in financial-domain articles

Hiroki Yamauchi2,1, 〇Koutarou Tamura1 (1.Uzabase, Inc., 2.Ehime Univ.)

Keywords:Named Entity Recognition, Natural Language Proessing

Named entity extraction task for organization names in news articles has achieved success in data for news wrote in general, natural text. However, we focused on articles in finance domain containing numerical information representing an organization like ticker, corporate id, securities code, and contents organized in a table structure. Previous studied rarely handled text including structured data, making this type of extraction difficult. In this study, we built text dataset containing these structured data mechanically and succeeded in improving our model performance with mixed data without degrading the accuracy of existing organization name extraction task.

Authentication for paper PDF access

A password is required to view paper PDFs. If you are a registered participant, please log on the site from Participant Log In.
You could view the PDF with entering the PDF viewing password bellow.

Password