Abstract | ||
---|---|---|
•Standard text classifiers are unable to capture clinical communication semantics.•Word embeddings, such as word2vec, are able to extract term relationships.•Convolutional neural networks (CNNs) can generate higher-order features for text.•Classifiers using CNNs and word embeddings improve communication classification.•The enhanced classifier outperforms standard methods by 1.5–10%. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1016/j.jbi.2017.08.014 | Journal of Biomedical Informatics |
Keywords | Field | DocType |
CNN,AUC,UMLS,CUI,STY,MHAV,VUMC,NLP,BoW,POS,NER,HER,SD,NLTK,DNN,RF,LR,ACC | Bag-of-words model,Data mining,Information retrieval,Convolutional neural network,Computer science,Patient portal,Word embedding,Word2vec,Random forest,Classifier (linguistics),Unified Medical Language System | Journal |
Volume | ISSN | Citations |
74 | 1532-0464 | 2 |
PageRank | References | Authors |
0.38 | 47 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Lina Sulieman | 1 | 2 | 2.07 |
David Gilmore | 2 | 8 | 0.88 |
Christi French | 3 | 2 | 0.38 |
Robert M Cronin | 4 | 63 | 9.71 |
Gretchen Purcell Jackson | 5 | 45 | 9.41 |
Matthew Russell | 6 | 35 | 3.15 |
Daniel Fabbri | 7 | 23 | 12.03 |