neo4j在线增量批量插入三元组

最新推荐文章于 2025-06-11 16:10:39 发布

原创

最新推荐文章于 2025-06-11 16:10:39 发布 · 2.4k 阅读

3 ·

CC 4.0 BY-SA版权

文章标签：

#neo4j #批量 #增量

本文介绍了如何使用Python生成csv文件，并详细步骤进行Neo4j数据库的节点与关系的批量及增量插入操作，强调了每种关系需单独导入的重要性。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

1.python代码生成words.csv、relations_daixiang.csv relations_yongxiang.csv relations_xiangguan.csv 注意：每种关系单独导入！

# 生成三元组
with open('d2019.txt', 'r', encoding='utf-8') as f:
    all = f.read()
records = all.split('*NEWRECORD')
tuples = []
mapd = {
   
   }
for record, i in zip(records[1:], range(1, len(records))):
    tuplesitem = []
    items = record.split('\n')
    idcount = 0
    for item in items:
        if item.startswith('MH = '):
            heading = re.findall('MH = ([^"]+)', item)[0]
            mapd[heading] = "D" + str(i)
        if item.startswith('ENTRY = '):
            entry = re.findall('[^\|]+', item)[0]
            entry = re.findall('ENTRY = ([^"]+)', entry)[0]
            tuples.append((heading, entry, '代项'))
            tuples.append((entry, heading, '用项'))
            if not mapd.get(entry,0):
                idcount = idcount + 1
                mapd[entry] = mapd[heading] + "_" + str(idcount)
        if item.startswith('PRINT ENTRY = '):
            printentry = re.findall('[^\|]+', item)[0]
            printentry = re.findall('PRINT ENTRY = ([^"]+)', printentry)[0]
            tuples.append((heading, printentry, '代项'))
            tuples.append((printentry, heading, '用项'))
            if not mapd.get(printentry, 0):
                idcount = idcount + 1
                mapd[printentry] = mapd[heading] + "_" + str(idcount)
        if item.startswith('MN = '):
            nid = re.findall('MN = ([^"]+)', item)[0]
            hypernyms = find_hypernym(nid)
            for hypernym in hypernyms:
                tuples.append((heading, hypernym, '上位词'))
            hyponyms = find_hyponym(nid)
            for hyponym in hyponyms:
                tuples.append((heading, hyponym, '下位词'))
            homoionyms = find_homoionym(nid)
            for homoionym in homoionyms:
                tuples.append((heading, homoionym, '兄弟'