def query_major_rc():
sql = 'SELECT TITLE,COUNT(TITLE),MOST_MONEY\
from job\
where info like "%软件测试%"\
group by title having count(*) > 0 \
ORDER BY COUNT(title) DESC\
LIMIT 10\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_list = []
res = list(cur.fetchall())
res_sorted_by_money = money_int.Money_Int(res)
for i in range(len(res_sorted_by_money)):
data_list.append({
'name': res_sorted_by_money[i][0],
'value': (int(res_sorted_by_money[i][1] / 1000 / 12))
})
print(jsonify(data_list))
return jsonify(data_list)
def Gensim_Search(text_url, keyword):
if os.path.exists(text_url):
pass
else:
text = ''
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
sql = 'select keyword from job'
cur.execute(sql)
for i in (cur.fetchall()):
text = text + str(i[0])
with open(text_url, 'w+', encoding='utf-8') as f:
f.write(text)
up_and_low.Up_And_Low(text_url)
sentences = word2vec.Text8Corpus(text_url)
model = word2vec.Word2Vec(sentences, size=200)
model.save('./textmodel.model')
try:
most_similar_word = model.most_similar(keyword.upper())
return most_similar_word
except:
return None
def query_talents():
sql = 'SELECT title,COUNT(title) as titlenum\
FROM job\
WHERE info LIKE "%大数据%" and title LIKE "%师"\
group by title having count(*) > 0\
ORDER BY COUNT(title) DESC\
LIMIT 10\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_dict = {}
data_dict['names'] = []
data_dict['values'] = []
data_dict['extra'] = []
res = list(cur.fetchall())
num = 0
for i in range(len(res)):
num += res[i][1]
for i in range(len(res)):
data_dict['names'].append({"name": res[i][0]})
data_dict['extra'].append(res[i][0])
for i in range(len(res)):
data_dict['values'].append(int((res[i][1])))
# print(data_dict)
return jsonify(data_dict)
def query_city():
sql = 'SELECT place_province,count(place_province) as num\
FROM job\
WHERE info LIKE "%大数据%"\
group by place_province having count(*) > 0\
ORDER BY COUNT(title) DESC\
LIMIT 10\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_list = []
res = list(cur.fetchall())
num = 0
for i in range(len(res)):
num += res[i][1]
for i in range(len(res)):
data_list.append({'name': res[i][0], 'value': int((res[i][1]))})
# print(data_dict)
return jsonify(data_list)
def query_demand_dmt():
sql = 'SELECT TITLE,COUNT(TITLE)\
from job\
where info like "%多媒体%"\
group by title having count(*) > 0\
ORDER BY COUNT(title) DESC\
LIMIT 10\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_list = []
res = list(cur.fetchall())
num = 0
for i in range(len(res)):
num += res[i][1]
for i in range(len(res)):
data_list.append({'name': res[i][0], 'value': int((res[i][1]))})
# print(data_dict)
return jsonify(data_list)
def query_salary_bj():
sql = 'SELECT place_city,least_money,most_money\
FROM job\
WHERE place_province="北京" AND place_city!="北京"\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_dict = {}
res = list(cur.fetchall())
is_not_same = []
res_sorted_by_least = money_least.Money_Least(res)
res_sorted_by_most = money_most.Money_Most(res)
res_fi1 = []
res_fi2 = []
for i in res_sorted_by_least:
if i["地区名"] not in is_not_same:
is_not_same.append(i["地区名"])
res_fi1.append({"地区名": i["地区名"], "最低工资": i["最低工资"]})
is_not_same = []
for i in res_sorted_by_most:
if i["地区名"] not in is_not_same:
is_not_same.append(i["地区名"])
res_fi2.append({"地区名": i["地区名"], "最高工资": i["最高工资"]})
for i in range(len(res_fi1)):
res_fi1[i]["最高工资"] = res_fi2[i]["最高工资"]
#{'numsh':[最高],'numsl':[最低],'names':[名称]}
data_dict['numsh'] = []
data_dict['numsl'] = []
data_dict['names'] = []
for i in range(len(res_fi1)):
data_dict['numsh'].append(res_fi1[i]['最高工资'])
data_dict['numsl'].append(res_fi1[i]['最低工资'])
data_dict['names'].append(res_fi1[i]['地区名'])
# print(data_dict)
data_dict['names'].remove('昌平区')
del data_dict['numsh'][5]
del data_dict['numsl'][5]
print(data_dict)
return jsonify(data_dict)
def query_location():
sql = 'SELECT DISTINCT place_province,count(place_province)\
FROM job\
group by place_province having count(*) > 0\
'
cur = db_connect.Db_Connect(host='localhost',
user='******',
passwd='1234',
db='51job',
port=3306,
charset='utf8')
cur.execute(sql)
data_list = []
res = list(cur.fetchall())
num = 0
for i in range(len(res)):
num += res[i][1]
for i in range(len(res)):
data_list.append({'name': res[i][0], 'value': int((res[i][1]))})
# print(data_dict)
return jsonify(data_list)
def Resume(place,major,text):
if text!='':
cur = db_connect.Db_Connect(host='localhost', user='******', passwd='1234', db='51job', port=3306, charset='utf8')
sql = "select distinct title,company,least_money,most_money,keyword,info from job where keyword like '%" + major + "%' or info like '%" + major + "%' or title like '%" + major + "%' and place_province like '%" + place +"%' or place_city like '%" + place +"%' limit 100"
cur.execute(sql)
#resume是添加交集功能，
data = cur.fetchall()
doc_test=text
#print(doc0,doc1,doc_test)
all_doc_chara=[]
all_doc=[]
try:
for i in data:
all_doc.append(i[0]+'-!-!-'+i[1]+'-!-!-'+i[2]+'-!-!-'+i[3]+'-!-!-'+i[4].split(' ')[5])
all_doc_chara.append("".join(re.sub(',|!|\?','',i[0].upper()))+"".join(re.sub(',|!|\?| ',' ',i[5].upper())))
except:
pass
all_doc_list=[]
for doc in all_doc_chara:
doc_list=[word for word in jieba.cut(doc)]
all_doc_list.append(doc_list)
#print(all_doc_list)#原文本的分词列表
doc_test_list=[word for word in jieba.cut(doc_test)]
#print(doc_test_list)#测试文本的分词列表
dictionary=corpora.Dictionary(all_doc_list)
#print(dictionary.keys())#原文本的字典键
#print(dictionary.token2id)#原文本键键名对应
corpus=[dictionary.doc2bow(doc) for doc in all_doc_list]
#print(corpus)#原文本键和出现次数[[(),()],[(),()]]
doc_test_vec=dictionary.doc2bow(doc_test_list)
#print(doc_test_vec)
tfidf=models.TfidfModel(corpus)#不同的转化需要不同的参数，在TF-IDF转化中，训练的过程就是简单的遍历训练语料库(corpus)，然后计算文档中每个特征的频率。
#找到有特征性的词作为区分的标准，tf计算是局部的，idf计算是全局的
#tf-idf算法是创建在这样一个假设之上的：对区别文档最有意义的词语应该是那些在文档中出现频率高，而在整个文档集合的其他文档中出现频率少的词语
#但是在本质上idf是一种试图抑制噪声的加权，并且单纯地认为文本频率小的单词就越重要，文本频率大的单词就越无用，显然这并不是完全正确的。idf的简单结构并不能有效地反映单词的重要程度和特征词的分布情况，使其无法很好地完成对权值调整的功能，所以tf-idf法的精度并不是很高。
index=similarities.SparseMatrixSimilarity(tfidf[corpus],num_features=len(dictionary.keys()))
sim=index[tfidf[doc_test_vec]]
# print(sim)
res=sorted(enumerate(sim),key=lambda item:-item[1])
res_list=[]
child_res_dict={}
child_res_dict['node']=[]
child_res_dict['link']=[]
print(res)
# print(len(res))
if len(res) <= 5:
cou_num = 1
for i in range(0,len(res)):
if res[i][1] > 0.005:
node_dict1 = {'name': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num + 1) * 10, 'y': 200,
'symbolSize': size1,
'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
"draggable": "true"}
node_dict2 = {'name': '序号:' + str(i) + ',职位名:' + all_doc[res[i][0]].split('-!-!-')[0] + ',工资区间:' + \
all_doc[res[i][0]].split('-!-!-')[2] + '-' + \
all_doc[res[i][0]].split('-!-!-')[3],
'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num + 1) * 10, 'y': 200,
'symbolSize': size2,
'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
"draggable": "true"}
# node_dict3 = {'name': '序号:' + str(i) + ',关键词:' + all_doc[res[i][0]].split('-!-!-')[4],
# 'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num + 1) * 10, 'y': 200,
# 'symbolSize': size2,
# 'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
# "draggable": "true"}
cou_num += 3
child_res_dict['node'].append(node_dict1)
child_res_dict['node'].append(node_dict2)
# child_res_dict['node'].append(node_dict3)
link_dict1 = {'source': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'name': '', 'value': '', 'label': '',
'lineStyle': {"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
link_dict2 = {'source': '序号:' + str(i) + ',职位名:' + all_doc[res[i][0]].split('-!-!-')[0] + ',工资区间:' + \
all_doc[res[i][0]].split('-!-!-')[2] + '-' + \
all_doc[res[i][0]].split('-!-!-')[3],
'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'name': '', 'value': '', 'label': '',
'lineStyle': {"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
# link_dict3 = {'source': '序号:' + str(i) + ',关键词:' + all_doc[res[i][0]].split('-!-!-')[4],
# 'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
# 'name': '', 'value': '', 'label': '',
# 'lineStyle': {"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
child_res_dict['link'].append(link_dict1)
child_res_dict['link'].append(link_dict2)
# child_res_dict['link'].append(link_dict3)
child_res_dict['node'].append({'name': '分析结果', 'value': '', 'x': 10, 'y': 200, 'symbolSize': size0,
'label': {'normal': {'position': 'inside', 'fontSize': 14, 'color': '#FF6633'}},
"draggable": "true"})
copy_list = child_res_dict['link'].copy()
for j in range(0, len(copy_list), 3):
child_res_dict['link'].append({'source': copy_list[j]['source'],
'target': child_res_dict['node'][len(child_res_dict['node']) - 1]['name'],
'name': '', 'value': '', 'label': '', 'lineStyle': {
"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}})
res_list.append(child_res_dict)
else:
cou_num = 1
for i in range(0,8):
if res[i][1] > 0.005:
node_dict1 = {'name': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num+1)*10, 'y': 200,
'symbolSize': size1,
'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
"draggable": "true"}
node_dict2 = {'name': '序号:' + str(i) + ',职位名:' + all_doc[res[i][0]].split('-!-!-')[0] + ',工资区间:' + \
all_doc[res[i][0]].split('-!-!-')[2] + '-' + \
all_doc[res[i][0]].split('-!-!-')[3],
'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num + 1) * 10, 'y': 200,
'symbolSize': size2,
'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
"draggable": "true"}
# node_dict3 = {'name': '序号:' + str(i) + ',关键词:' + all_doc[res[i][0]].split('-!-!-')[4],
# 'value': '%.2f' % (res[i][1] * 1000), 'x': (cou_num + 1) * 10, 'y': 200,
# 'symbolSize': size2,
# 'label': {'normal': {'position': 'inside', 'fontSize': 10, 'color': '#FF6633'}},
# "draggable": "true"}
cou_num += 3
child_res_dict['node'].append(node_dict1)
child_res_dict['node'].append(node_dict2)
# child_res_dict['node'].append(node_dict3)
link_dict1 = {'source':'序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'name': '', 'value': '', 'label': '','lineStyle':{"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
link_dict2 = {'source': '序号:' + str(i) + ',职位名:' + all_doc[res[i][0]].split('-!-!-')[0] + ',工资区间:' + \
all_doc[res[i][0]].split('-!-!-')[2] + '-' + \
all_doc[res[i][0]].split('-!-!-')[3],
'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'name': '', 'value': '', 'label': '','lineStyle':{"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
link_dict3 = {'source': '序号:' + str(i) + ',关键词:' + all_doc[res[i][0]].split('-!-!-')[4],
'target': '序号:' + str(i) + ',公司名:' + all_doc[res[i][0]].split('-!-!-')[1],
'name': '', 'value': '', 'label': '','lineStyle':{"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}}
child_res_dict['link'].append(link_dict1)
child_res_dict['link'].append(link_dict2)
child_res_dict['link'].append(link_dict3)
child_res_dict['node'].append({'name': '分析结果', 'value': '', 'x': 10, 'y': 200, 'symbolSize': size0,
'label': {'normal': {'position': 'inside', 'fontSize': 14, 'color': '#FF6633'}},
"draggable": "true"})
copy_list = child_res_dict['link'].copy()
for j in range(0, len(copy_list), 3):
child_res_dict['link'].append({'source': copy_list[j]['source'],
'target': child_res_dict['node'][len(child_res_dict['node']) - 1]['name'],
'name': '', 'value': '', 'label': '','lineStyle':{"normal": {"width": 2.0, "curveness": 0.2, "color": '#FF6633'}}})
res_list.append(child_res_dict)
return res_list
else:
return []
