ロード中...
Facilitating information extraction without annotated data using unsupervised and positive-unlabeled learning
Information extraction (IE), the distillation of specific information from unstructured data, is a core task in natural language processing. For rare entities (<1% prevalence), collection of positive examples required to train a model may require an infeasibly large sample of mostly negative ones...
保存先:
| 出版年: | AMIA Annu Symp Proc |
|---|---|
| 主要な著者: | , , , , |
| フォーマット: | Artigo |
| 言語: | Inglês |
| 出版事項: |
American Medical Informatics Association
2021
|
| 主題: | |
| オンライン･アクセス: | https://ncbi.nlm.nih.gov/pmc/articles/PMC8075513/ https://ncbi.nlm.nih.gov/pubmed/33936440 |
| タグ: |
タグ追加
タグなし, このレコードへの初めてのタグを付けませんか!
|