SDU创新实训5. 数据库操作功能模块实现（CQL语句配合py2neo）

1. 根据关键词查询每年该种类关键词发布的数量；（发表年度趋势）
   match (:Keyword{name:‘人工智能’})-[:is_Keyword_to]->(n:Paper)
   return n.year,count(*)
   
2. 关键词共现网络
   这个是预备工作

先进行初始化
LOAD CSV WITH HEADERS FROM "file:///keywordNet.csv" AS line
MATCH (entity1:Keyword{name:line.name}),(entity2:Keyword{name:line.name2})
where not  (entity1)-[:co_occurrence]->(entity2) 
CREATE (entity1)-[:co_occurrence{count:0}]->(entity2)
再次加载，更改共现次数
LOAD CSV WITH HEADERS FROM "file:///keywordNet.csv" AS line
MATCH (entity1:Keyword{name:line.name})-[r:co_occurrence]->(entity2:Keyword{name:line.name2})
SET r.count=r.count+1

LOAD CSV WITH HEADERS FROM "file:///authorNet.csv" AS line
MATCH (entity1:Author{name:line.name}),(entity2:Author{name:line.name2})
where not  (entity1)-[:cooperative]->(entity2) 
CREATE (entity1)-[:cooperative{count:0}]->(entity2)

LOAD CSV WITH HEADERS FROM "file:///authorNet.csv" AS line
MATCH (entity1:Author{name:line.name})-[r:cooperative]->(entity2:Author{name:line.name2})
SET r.count=r.count+1

返回作者共现网络

MATCH (n1:Author{name:'任婷婷'})-[r1:cooperative]-(n2:Author) 
 with n1,r1,n2
  MATCH (n2)-[r2: cooperative]-(n3:Author)
  where n3.name <>n1.name
  RETURN n1,r1,n2,r2,n3,id(n1),id(n2),id(n3)

返回关键词共现网络

MATCH (n1:Keyword{name:"拉普拉斯特征映射"})-[r1:co_occurrence]-(n2:Keyword)
 with n1,r1,n2
  MATCH (n2)-[r2:co_occurrence]-(n3:Keyword)
where n3.name <>n1.name
 RETURN n1,r1,n2,r2,n3,id(n1),id(n2),id(n3)

3. 作者分析器
   发文量：

match (:Author{name:'钟义信'})-[:is_Author]->(n:Paper)
return n.year,count(*)
order by n.year

下载量,引用量

match (:Author{name:'钟义信'})-[:is_Author]->(n:Paper)
return n.title,n.download,n.quote,n.year
order by n.year

关键词(作者发布的文章关键词频次）

match (:Author{name:'钟义信'})-[:is_Author]->(n:Paper)
with n
match (n)-[:Keyword]->(k:Keyword)
with k.name as key
return key,count(*) as count
order by count desc

6. 某类文章关键词共现排名
   根据关键词共现网络得到和该关键词共同出现的关键词

match (:Keyword{name:'人工智能'})-[r:co_occurrence]-(k:Keyword)
return k.name,r.count
order by r.count desc

6. 某类文章文献来源分布

每个source的对应关键词的发布文章数量

MATCH (k:Keyword{name:'人工智能'})-[r:is_Keyword]->(n:Paper)
return n.source,count(*) as c
order by c desc

根据srcDatabase

MATCH (k:Keyword{name:'人工智能'})-[r:is_Keyword]->(n:Paper)
return n.srcDatabase,count(*) as c
order by c desc

可视化：

8. 某类文章作者分布
   这个author有的是单独有的是好几个，后端用NLP处理 ，先不搞
   MATCH (k:Keyword{name:‘人工智能’})-[r:is_Keyword]->(n:Paper)
   return n.author,count(*) as c
   order by c desc