Deep learning with word embeddings improves biomedical named entity recognition.
Deep learning with word embeddings improves biomedical named entity recognition.
复制标题
DOI：
10.1093/bioinformatics/btx228
复制
发表时间：
2017-07-15
期刊：
影响因子：
--
通讯作者：
Leser U
中科院分区：
文献类型：
--
作者：
Habibi M;Weber L;Neves M;Wiegandt DL;Leser U
文本挖掘已经成为生物医学研究的重要工具。最基本的文本挖掘任务是识别生物医学命名实体（NER），如基因，化学物质和疾病。目前的NER方法依赖于预定义的功能，试图捕捉特定的表面属性的实体类型，典型的本地上下文，背景知识和语言信息的属性。最先进的工具是特定于实体的，因为字典和经验最佳特征集在实体类型之间不同，这使得它们的开发成本很高。此外，特征通常针对特定的黄金标准语料库进行优化，这使得质量度量的外推变得困难。我们证明了一种基于深度学习和统计词嵌入的完全通用的方法[称为长短期记忆网络条件随机场（LSTM-CRF）]优于最先进的实体特定NER工具，并且通常是大幅度的。为此，我们比较了LSTM-CRF在33个数据集上的性能，这些数据集覆盖了5个不同的实体类，并与一流的NER工具和实体不可知的CRF实现进行了比较。平均而言，LSTM-CRF的F1分数比基线高5%，主要是由于回忆的急剧增加。LSTM-CRF的源代码可在https://github.com/glample/tagger上获得，语料库的链接可在https://corposaurus.github.io/corpora/上获得。
Text mining has become an important tool for biomedical research. The most fundamental text-mining task is the recognition of biomedical named entities (NER), such as genes, chemicals and diseases. Current NER methods rely on pre-defined features which try to capture the specific surface properties of entity types, properties of the typical local context, background knowledge, and linguistic information. State-of-the-art tools are entity-specific, as dictionaries and empirically optimal feature sets differ between entity types, which makes their development costly. Furthermore, features are often optimized for a specific gold standard corpus, which makes extrapolation of quality measures difficult. We show that a completely generic method based on deep learning and statistical word embeddings [called long short-term memory network-conditional random field (LSTM-CRF)] outperforms state-of-the-art entity-specific NER tools, and often by a large margin. To this end, we compared the performance of LSTM-CRF on 33 data sets covering five different entity classes with that of best-of-class NER tools and an entity-agnostic CRF implementation. On average, F1-score of LSTM-CRF is 5% above that of the baselines, mostly due to a sharp increase in recall. The source code for LSTM-CRF is available at https://github.com/glample/tagger and the links to the corpora are available at https://corposaurus.github.io/corpora/.
登录
查看更多内容
影响因子：
8.6
作者：
Habibi M;Wiegandt DL;Schmedding F;Leser U
通讯作者：
Leser U
影响因子：
8.6
作者：
Batista-Navarro R;Rak R;Ananiadou S
通讯作者：
Ananiadou S
影响因子：
4.5
作者：
Dogan, Rezarta Islamaj;Leaman, Robert;Lu, Zhiyong
通讯作者：
Lu, Zhiyong
影响因子：
3.7
作者：
Akhondi SA;Klenner AG;Tyrchan C;Manchala AK;Boppana K;Lowe D;Zimmermann M;Jagarlapudi SA;Sayle R;Kors JA;Muresan S
通讯作者：
Muresan S
影响因子：
7.8
作者：
Graves, A;Schmidhuber, J
通讯作者：
Schmidhuber, J