Zhongda Xie 2015-01-12

来自cslt Wiki
2015年1月13日 (二) 13:31Xiezd讨论 | 贡献的版本

(差异) ←上一版本 | 最后版本 (差异) | 下一版本→ (差异)
跳转至: 导航搜索

Progress in Last Week:

Use biolerpipe for web document extraction.

Use the knowledge base UMLS, and map the whole document content into their corresponding CUIs(Concept Unique Identifiers).

Plan for this Week:

For every pair of concepts, dynamically compute the weight of their possible relations.

Consider the weight of relations in concept pair assessment, modify the InQuery string.