Biomedical Semantic Embeddings: Using Hybrid Sentences to Construct Biomedical Word Embeddings and its Applications
Description:
Word embeddings is a useful method that has shown enormous success in various NLP tasks, not only in open domain but also in biomedical domain. The biomedical domain provides various domain specific resources and tools that can be exploited to improve performance of these word embeddings. However, most of the research related to word embeddings in biomedical domain focuses on analysis of model architecture, hyper-parameters and input text. In this paper, we use SemMedDB to design new sentences …
more
Date:
December 2019
Creator:
Shaik, Arshad
Partner:
UNT Libraries