Papers by Apostolos Kemos

1 papers
Neural Semi-Markov Conditional Random Fields for Robust Character-Based Part-of-Speech Tagging (N19-1)

Copied to clipboard

Challenge: Character-level models of tokens are effective at dealing with within-token noise and out-of-vocabulary words.
Approach: They propose to eliminate the need for tokenizers by using a character-level semi-Markov conditional random field that uses neural networks for its character and segment representations.
Outcome: The proposed model outperforms state-of-the-art part-of speech taggers on a noisy English dataset.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations