MaterialBERT for Natural Language Processing of Materials Science Texts

Michiko Yoshitake; Fumitaka Sato; Hiroyuki Kawano; Hiroshi Teraoka

Presentation information

Oral presentation

23 Joint Session N "Informatics" » 23.1 Joint Session N "Informatics"

[19p-Z32-1~15] 23.1 Joint Session N "Informatics"

Fri. Mar 19, 2021 1:30 PM - 5:45 PM Z32 (Z32)

Tetsuhiko Miyadera(AIST), Tatsuya Yokoi(Nagoya Univ.), Yukinori Koyama(NIMS)

5:15 PM - 5:30 PM

[19p-Z32-14] MaterialBERT for Natural Language Processing of Materials Science Texts

〇Michiko Yoshitake¹, Fumitaka Sato^1,2, Hiroyuki Kawano^1,2, Hiroshi Teraoka^1,2 (1.NIMS, 2.RidgeLinez)

Keywords:text mining, natural language processing, BERT

We developed pre-training model of BERT using materials science scientific papers as a copus (we call MaterialBERT) for ulitization to natural language processing of materials science texts.