KG知识图谱

KG知识图谱

1

知识图谱创建过程：major steps of an overall process model:
1 knowledge creation,
2 knowledge hosting,
3 knowledge curation, and
4 knowledge deployment

-

1. Knowledge Creation

knowledge acquisition/ knowledge engineering

-Automatic Annotation Tools

Automatic annotation tools extract data from the web using natural language processing (NLP)
and machine learning (ML)

There are many tools or libraries available, such as GATE22 for text analysis and language
processing, OpenNLP23 that supports the most common NLP tasks, and RapidMiner
-

2. Knowledge Hosting
Collection, Storage, and Retrieval of Knowledge Graphs

3. Knowledge Curation

knowledge assessment, cleaning, and enrichment

4. Knowledge Deployment/Application
-

_XAI

-

-

Ref:

Baidu hugegraph benchmark:

https://hugegraph.github.io/hugegraph-doc/performance/hugegraph-benchmark-0.5.6.html

【知识图谱】百度大规模知识图谱构建及智能应用

https://mp.weixin.qq.com/s/PvLrrldXXt8UQHTCpImzdw

宋勋超，硕士毕业于浙江大学，百度知识图谱部主任研发架构师。参与了百度知识图谱设计及构建的整体流程，具有丰富的知识图谱实践经验。目前主要负责通用知识图谱构建、语义理解、图谱架构等技术，同时负责百度行业知识图谱相关工作。研发成果在百度搜索、信息流、DuerOS，行业图谱产品等多项产品中广泛应用。

百度知识图谱技术及应用

https://mp.weixin.qq.com/s/wmPnW8N6NvuQCNzEAUJ4PQ

分享嘉宾：王泉博士 百度 资深研发工程师

项目实战：如何构建知识图谱

https://blog.youkuaiyun.com/qq_35273499/article/details/80259821

知识图谱构建过程

https://blog.youkuaiyun.com/u013378306/article/details/105970931

[Github]知识图谱的应用示例

https://zhuanlan.zhihu.com/p/77339370

知识图谱-给AI装个大脑 https://zhuanlan.zhihu.com/knowledgegraph
知识图谱综合,2005 年至 2019 年 80 篇知识图谱领域经典论文集
  https://blog.youkuaiyun.com/weixin_43269174/article/details/100168900
知识图谱架构（Knowledge Graph） https://zyc88.blog.youkuaiyun.com/article/details/105367041
案例：
基于知识图谱的问答系统(KBQA)  https://blog.youkuaiyun.com/keyue123/article/details/85266355
基于电影知识图谱的智能问答系统 https://blog.youkuaiyun.com/appleyk/category_7667380.html

关键技术：1,实体识别
【NLP-NER】命名实体识别详解之一, https://zhuanlan.zhihu.com/p/88544122
命名实体识别-BiLSTM+CRF,  https://zhuanlan.zhihu.com/p/105282809
手撕 BiLSTM-CRF   https://zhuanlan.zhihu.com/p/97676647

BERT + BiLSTM + CRF 命名实体识别, https://zhuanlan.zhihu.com/p/351247764

知识抽取-实体及关系抽取 https://zhuanlan.zhihu.com/p/44772023
知识抽取-事件抽取           https://zhuanlan.zhihu.com/p/50903358

关键技术：2,关系抽取(分类)

西多士NLP,信息抽取——关系抽取，，
https://www.cnblogs.com/sandwichnlp/p/12020066.html

关系抽取(分类)总结，http://shomy.top/2018/02/28/relation-extraction/

关系抽取(分类)总结， https://zhuanlan.zhihu.com/p/44772023
nlp中的实体关系抽取方法总结  https://zhuanlan.zhihu.com/p/77868938
https://blog.youkuaiyun.com/qq_27590277/article/details/107133347

概率图模型体系：HMM、CRF  https://zhuanlan.zhihu.com/p/33397147

关键技术：3,消歧

======

NELL（never-ending language learner）的framework,,
https://blog.youkuaiyun.com/john159151/article/details/53573441