knowledge-graph-learning
knowledge-graph-learning copied to clipboard
BrambleXu
A curated list of awesome knowledge graph tutorials, projects and communities.
**Summary:** 将BERT应用到科学期刊领域的研究 **Resource:** - [pdf](https://aclanthology.org/D19-1371/) - [code](https://github.com/allenai/scibert/) - [paper-with-code]( **Paper information:** - Author: - Dataset: [ACL ARC](https://paperswithcode.com/dataset/acl-arc-1), [BC5CDR](https://paperswithcode.com/dataset/bc5cdr), [ChemProt](https://paperswithcode.com/dataset/chemprot), [EBM-NLP](https://paperswithcode.com/dataset/ebm-nlp), [GENIA](https://paperswithcode.com/dataset/genia), [JNLPBA](https://paperswithcode.com/dataset/jnlpba), [Microsoft Academic Graph](https://paperswithcode.com/dataset/microsoft-academic-graph), [NCBI Disease](https://paperswithcode.com/dataset/ncbi-disease-1), [Paper Field](https://paperswithcode.com/dataset/paper-field), [Pubmed](https://paperswithcode.com/dataset/pubmed),...
**Summary:** 给BERT的预训练又添加了两个额外的任务，一个是句子内用于学习单词顺序的任务，一个是不同句子间用于学习句子间顺序的任务。 **Resource:** - [pdf](https://openreview.net/pdf?id=BJgQ4lSFPH) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** 从设计上来说，这个模型对于相似度的贡献应该比较小 **Next Reading:**
**Summary:** 论文有52页，有点长，对于多语言相似度的研究，挺有参考价值的。呃，没有日语。。 本研究创建了一个新的数据集，构建了不同语言的相似度单词组 **Resource:** - [pdf](https://aclanthology.org/2020.cl-4.5/) - [Download Multi-SimLex](https://multisimlex.com/#download) - [paper-with-code]( **Paper information:** - Author: - Dataset: [CARD-660](https://paperswithcode.com/dataset/card-660), [MLQA](https://paperswithcode.com/dataset/mlqa), [Panlex](https://paperswithcode.com/dataset/panlex), [TyDi QA](https://paperswithcode.com/dataset/tydi-qa), [XNLI](https://paperswithcode.com/dataset/xnli), [XQuAD](https://paperswithcode.com/dataset/xquad) - keywords: **Notes:** **Model Graph:** **Result:**：...
**Summary:** PolyglotIE，一个基于web的多语言IE工具。 **Resource:** - [pdf](https://aclanthology.org/C16-2056/) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** **Next Reading:**
**Summary:** 利用transfer learning将英语中学到的内容给日语学习。 **Resource:** - [pdf](https://aclanthology.org/N19-2023/) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: [CoNLL-2003](https://paperswithcode.com/dataset/conll-2003) - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** 中文其实更适合做source，只可惜中文能用的数据太少，这方面确实不如英语 **Next Reading:**
**Summary:** 针对临床文本，进行NER和RE的验证 **Resource:** - [pdf](https://www.jstage.jst.go.jp/article/pjsai/JSAI2022/0/JSAI2022_1J4OS13a03/_pdf) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: - keywords: **Notes:** 没找到数据集，但是这数量也太少了吧，有参考意义吗。。 **Model Graph:** **Result:**： **Thoughts:** **Next Reading:**
**Summary:** 针对临床文本的entity-relation抽取，构建了一个系统。 **Resource:** - [pdf](https://proceedings-of-deim.github.io/DEIM2022/papers/A34-2.pdf) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: EMR的数据集 - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** **Next Reading:**
**Summary:** 从履历中提取信息。 **Resource:** - [pdf](https://anlp.jp/proceedings/annual_meeting/2018/pdf_dir/P10-5.pdf) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** 偏应用的研究，主要是这个应用领域比较偏 **Next Reading:**
**Summary:** 针对日语医疗信息抽取。剔除了一种新的关系标注架构，用于利用entity之间的医疗和时间信息。 **Resource:** - [pdf](https://aclanthology.org/2022.lrec-1.397/) - [code]( - [paper-with-code]( **Paper information:** - Author: - Dataset: - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** **Next Reading:**
**Summary:** 针对few-shot，提出了一种新的relation represention. **Resource:** - [pdf](https://aclanthology.org/2022.findings-acl.62/) - lylylylylyly/simplefsre - [paper-with-code]( **Paper information:** - Author: - Dataset: [FewRel](https://paperswithcode.com/dataset/fewrel) - keywords: **Notes:** **Model Graph:** **Result:**： **Thoughts:** **Next Reading:**