TF-IDF提取关键词

　　今天完成了提取字符串关键词的这一部分，代码如下：

# -*- coding: gbk -*-
import jieba.analyse
import pymysql


# 打开连接
def open_conn(dbname):
    db = pymysql.connect(
        host="localhost",
        port=3306,
        user="root",
        passwd="******",
        db=dbname,
        charset="utf8")
    return db

# 遍历查询
def query(db):
    cursor = db.cursor()
    sql = "select ID,key_words,result_summary from summary_table"
    cursor.execute(sql)
    for each in cursor.fetchall():
        ID = each[0]
        key_words = each[1]
        result_summary = each[2]
        if key_words is None:                  
            if len(result_summary) > 10:
                new_key_word = get_keyword_by_TFIDF(result_summary)
                update(db,ID,new_key_word)
                print(ID,new_key_word)

# 修改数据  参数（db,ID,关键词）
def update(db,ID,keyword):
    cursor = db.cursor()
    sql = " update summary_table set key_words = %s where ID = %s"
    cursor.execute(sql,(keyword,ID))
    db.commit()

##词性：n:名词,v:动词，ns:地名，vn:名动词，nt:机构团体
# TF-IDF算法提取关键词
def get_keyword_by_TFIDF(result_summary):
    keywords = " ".join(jieba.analyse.extract_tags(result_summary, topK=5, withWeight=False, allowPOS=(["ns", "n", "vn", "v","nt"])))
    return keywords
#TextRank算法
def get_keyword_by_TextRank(result_summary):
    keywords = " ".join(jieba.analyse.textrank(result_summary, topK=5, withWeight=False, allowPOS=("ns", "n", "vn", "v","nt")))
    return keywords


if __name__ == '__main__':
    query(open_conn("datax"))