https://github.com/zhpmatrix/tianchi-ruijin
瑞金医院MMC人工智能辅助构建知识图谱大赛-baseline
https://github.com/zhpmatrix/tianchi-ruijin
knowledge-graph mmc ner tianchi
Last synced: 17 days ago
JSON representation
瑞金医院MMC人工智能辅助构建知识图谱大赛-baseline
- Host: GitHub
- URL: https://github.com/zhpmatrix/tianchi-ruijin
- Owner: zhpmatrix
- Created: 2018-10-15T09:10:11.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2018-10-30T13:51:32.000Z (over 6 years ago)
- Last Synced: 2025-03-26T17:11:19.485Z (about 1 month ago)
- Topics: knowledge-graph, mmc, ner, tianchi
- Language: Python
- Size: 3.87 MB
- Stars: 67
- Watchers: 7
- Forks: 29
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
#### 介绍
[瑞金医院MMC人工智能辅助构建知识图谱大赛-第一赛季](https://tianchi.aliyun.com/competition/introduction.htm?spm=5176.100066.0.0.55d733afYIca7m&raceId=231687)
#### 数据
从给定数据中抽出一份作为测试样本,对余下样本按照train : dev = 10 : 1的比例划分
#### 模型
BiLSTM+CRF
#### 实现
Linux+Python3.6+Keras2.2.4+Tensorflow1.11.0
#### 结果
线上分数=0.71
#### 思考
赛题第一个阶段建模为一个命名实体识别任务(NER),但是还没思考好第二阶段的方法。通过调研发现,实体识别和关系抽取两个子任务可以通过一个模型实现,比起前者,后者似乎更有趣一些。