使用OpenIE提取给定实体的关系

cd stanford-corenlp-full-<version>
java -mx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer -port 9000 -timeout 15000

import requests
import json

def extract_relations(text, entity):
    url = 'http://localhost:9000'
    params = {
        'annotators': 'openie',
        'outputFormat': 'json'
    }
    data = {
        'text': text
    }
    response = requests.post(url, params=params, data=data)
    result = response.json()

    relations = []
    for sentence in result['sentences']:
        for triple in sentence['openie']:
            if entity in triple['subject'] or entity in triple['object']:
                relations.append(triple)
    
    return relations

# 示例文本
text = "Barack Obama was born in Hawaii. He was elected president in 2008. Obama served as the 44th president of the United States."

# 给定实体
entity = "Obama"

# 提取关系
relations = extract_relations(text, entity)

# 打印结果
for relation in relations:
    print(f"Subject: {relation['subject']}")
    print(f"Relation: {relation['relation']}")
    print(f"Object: {relation['object']}")
    print()

Barack Obama was born in Hawaii. He was elected president in 2008. Obama served as the 44th president of the United States.

Subject: Obama
Relation: served as
Object: the 44th president of the United States

Subject: He
Relation: was elected
Object: president