entity-linking

实体链接(命名实体消歧)

目标: 针对document或者query中的名实体(name mention)确定其具体的含义

1.getcanbywiki:使用维基百科API返回与输入字符串相似度高的entity
2.getcanbyRule: 采用基于搜索规则的方法生成候选实体集
3.getcanbyPrior：采用基于先验概率的方法生成候选实体集

1.每两个候选实体之间构建关联对
2.使用DFS在DBpedia中探索关联对之间的路径
3.所有路径与实体构成graph

1.利用PageRank计算图节点(代表实体)的值
2.选择得分最高的作为该name mention对应的知识库中的链接实体

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
dataset		dataset
KBPparser.py		KBPparser.py
PriorRuleRoad1Result.txt		PriorRuleRoad1Result.txt
PriorRuleRoadResult.txt		PriorRuleRoadResult.txt
README.md		README.md
RuleRoad1Result.txt		RuleRoad1Result.txt
RuleRoadResult.txt		RuleRoadResult.txt
booleantest.txt		booleantest.txt
entitylink.py		entitylink.py
entitylink.pyc		entitylink.pyc
getPairArrayPath1.txt		getPairArrayPath1.txt
getcan.py		getcan.py
getcan.pyc		getcan.pyc
getcanbyfile.py		getcanbyfile.py
getcanbywiki.py		getcanbywiki.py
getcanbywiki.pyc		getcanbywiki.pyc
getpath_new.py		getpath_new.py
graphconstrict.py		graphconstrict.py
labels_process.py		labels_process.py
ls.txt		ls.txt
natural.txt		natural.txt
new.txt		new.txt
new_1.txt		new_1.txt
pagelinks_process.py		pagelinks_process.py
pagerank.py		pagerank.py
parser.py		parser.py
pathfile1.txt		pathfile1.txt
screen.log		screen.log
test.log		test.log
test.py		test.py
test1.py		test1.py
test2.py		test2.py
typescript		typescript
wiki.txt		wiki.txt
wikiresult.txt		wikiresult.txt