Neural Network Based Named Entity Recognition

Speaker:
Jana Straková
Abstract:
We present the recent artificial neural network based methods for named entity recognition (NER, the identification and classification of proper names in  text). We briefly introduce the task and data, and quickly skimming through related work (Ratinov and Roth, 2009), we focus on recent advances in the field of artificial neural networks - word embeddings and character-level embeddings. Previous knowledge of these concepts is not necessary, we explain them in the context of NER. We present both the current English (Lample et al., 2016) and Czech (Strakova et al., 2016) state of the art.
Length:
00:58:43
Date:
06/03/2017
views:

Images:
Attachments: (video, slides, etc.)
81 MB
20 downloads
749 MB
21 downloads
101 MB
22 downloads
144 MB
20 downloads
296 MB
26 downloads