def extract_keyword_by_thulac(self):
sents = []
comm_list = self.dao.get_hotel_comments()
# 从语料中读取每一行并切分成子句
for comm in comm_list:
sents.extend(normal.get_sentences(comm[2]))
print "length of sentences:%d"%len(sents)
# 每个子句进行词性判读
pos_sents = []
for sent in sents:
try:
pos_sents.append(map(lambda x: x.split("_"), self.thu.cut(sent.encode("utf-8"))))
except:
print sent
continue
print "length of pos_sents:%d"%len(pos_sents)
# 分拣出名词,并进行统计
print "counting"
noun_dict = {}
for pos_sent in pos_sents:
for word in pos_sent:
if word[1] == "n":
if word[0] not in noun_dict:
noun_dict[word[0]] = 1
else:
noun_dict[word[0]] = noun_dict[word[0]] + 1
a = sorted(noun_dict.iteritems(),key=lambda asd:asd[1],reverse=True)
return a
def extract_keyword_by_thulac(self):
sents = []
comm_list = self.dao.get_hotel_comments()
# 从语料中读取每一行并切分成子句
for comm in comm_list:
sents.extend(normal.get_sentences(comm[2]))
print "length of sentences:%d" % len(sents)
# 每个子句进行词性判读
pos_sents = []
for sent in sents:
try:
pos_sents.append(
map(lambda x: x.split("_"),
self.thu.cut(sent.encode("utf-8"))))
except:
print sent
continue
print "length of pos_sents:%d" % len(pos_sents)
# 分拣出名词,并进行统计
print "counting"
noun_dict = {}
for pos_sent in pos_sents:
for word in pos_sent:
if word[1] == "n":
if word[0] not in noun_dict:
noun_dict[word[0]] = 1
else:
noun_dict[word[0]] = noun_dict[word[0]] + 1
a = sorted(noun_dict.iteritems(), key=lambda asd: asd[1], reverse=True)
return a
def main():
t = normal.zh2hans(text)
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
rank = textrank.TextRank(doc)
rank.solve()
for index in rank.top_index(5):
print(sents[index])
keyword_rank = textrank.KeywordTextRank(doc)
keyword_rank.solve()
for w in keyword_rank.top_index(5):
print(w)
def parse_keyword(text):
t = normal.zh2hans(text.decode("UTF-8"))
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
keywords = []
keyword_rank = textrank.KeywordTextRank(doc)
keyword_rank.solve()
for w in keyword_rank.top_index(5):
keywords.append(w)
return keywords
def query_xym():
offset = request.values['offset']
pagesize = request.values['pageSize']
resp = XymTripModel.query.offset(offset).limit(pagesize)
t = normal.zh2hans(
"随着智能手机和平板电脑的普及，相机也变得无处不在，而且分享照片也越来越简单。MOOC的明星教授说，把45分钟的讲座变成10分钟一段的视频让他们被迫“升级课程”。不是每个老师都能通过这种方式吸引一批学生，但是他们可以参考这个经验，为课堂制作自己的视频，例如实地考察录像。让整个班都出去跑一趟可能不可行，但利用视频和照片，可以把考察点“带”到课室中来。利用智能手机耳机上配备的话筒，还可以为视频配上讲解，从而高效地用多个视频介绍完一个知识点。"
)
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
rank = textrank.TextRank(doc)
rank.solve()
for index in rank.top_index(5):
print(sents[index])
keyword_rank = textrank.KeywordTextRank(doc)
keyword_rank.solve()
for w in keyword_rank.top_index(5):
print(w)
return "\'..."
def extract_keyword(self):
sents = []
comm_list = self.dao.get_hotel_comments()
# 从语料中读取每一行并切分成子句
for comm in comm_list:
sents.extend(normal.get_sentences(comm[2]))
print "length of sentences:%d" % len(sents)
# 每个子句进行词性判读
pos_sents = []
for sent in sents:
pos_sents.append(pseg.cut(sent))
print "length of pos_sents:%d" % len(pos_sents)
# 分拣出名词,并进行统计
print "counting"
noun_dict = {}
for pos_sent in pos_sents:
for key, type in pos_sent:
if type == "n":
if key not in noun_dict:
noun_dict[key] = 1
else:
noun_dict[key] = noun_dict[key] + 1
a = sorted(noun_dict.iteritems(), key=lambda asd: asd[1], reverse=True)
return a
def extract_keyword(self):
sents = []
comm_list = self.dao.get_hotel_comments()
# 从语料中读取每一行并切分成子句
for comm in comm_list:
sents.extend(normal.get_sentences(comm[2]))
print "length of sentences:%d"%len(sents)
# 每个子句进行词性判读
pos_sents = []
for sent in sents:
pos_sents.append(pseg.cut(sent))
print "length of pos_sents:%d"%len(pos_sents)
# 分拣出名词,并进行统计
print "counting"
noun_dict = {}
for pos_sent in pos_sents:
for key,type in pos_sent:
if type == "n":
if key not in noun_dict:
noun_dict[key] = 1
else:
noun_dict[key] = noun_dict[key] + 1
a = sorted(noun_dict.iteritems(),key=lambda asd:asd[1],reverse=True)
return a
def summaryOfSents(docs, limit=5):
'''
docs = [
'第一篇微博',
'第二篇微博',
...
]
'''
merged_doc = []
sents = []
for doc in docs:
sents_t = normal.get_sentences(doc)
sents += sents_t
for sent in sents_t:
words = jieba.lcut(sent)
words = normal.filter_stop(words)
merged_doc.append(words)
rank = textrank.TextRank(merged_doc)
rank.solve()
ret = []
for index in rank.top_index(limit):
ret.append(sents[index])
return ret
MOOC是学校的一种新形式，欧伯恩建议在起步的时候，先为每门课程的课件加上指针，再利用软件工具，就可以轻松根据学生的学习进度添加课程。他希望，在学生使用在线社区的同时，教师也能发现参与在线社区的方式。
5．从线上到线下
MOOC的一个缺陷就是无法组建高效的学习小组，而教师在这方面可以大有作为。当学生们看到其他同学更新了课程内容，他们就知道谁掌握了所学的知识，从而邀请这些同学合作完成任务，或向他们请教。我经常向教师们介绍这个例子：我在Google+圈子里发了一条信息，例如“明天我们会讨论矛盾冲突在吸引读者注意力方面的作用。今晚，在你回家的路上，拍一张照片或一段录像。用文字介绍你的见闻，以证明这个观点，并邀请其他同学参与讨论。”我收到的作业包括交通堵塞，猫狗对峙，被泡在水里的花园以及足球训练中的射门。第二天，学生们就可以归纳整理前一天晚上在网络上收集到的评论了。
6．用好你的相机
随着智能手机和平板电脑的普及，相机也变得无处不在，而且分享照片也越来越简单。MOOC的明星教授说，把45分钟的讲座变成10分钟一段的视频让他们被迫“升级课程”。不是每个老师都能通过这种方式吸引一批学生，但是他们可以参考这个经验，为课堂制作自己的视频，例如实地考察录像。让整个班都出去跑一趟可能不可行，但利用视频和照片，可以把考察点“带”到课室中来。利用智能手机耳机上配备的话筒，还可以为视频配上讲解，从而高效地用多个视频介绍完一个知识点。
将MOOC应用到传统课堂教学
随着大规模网络公开课的发展，教师可以考虑把在线教育的方法应用到自己的课堂教学中。MOOC的课程制作涉及比较复杂的技术，但使用这些课程几乎不费吹灰之力，而且成本也远远不及课程制作。没有加入edX或Coursera的大部分学校可以进行更多自创内容的尝试，就像自出版一样，这也是许多cMOOC的尝试。教师也可以向自己的目标努力。通过打开课堂，建立网络社区和制作教学视频，可以让更多的教师和学生享受到MOOC的投入带来的收益。
'''
from snownlp import normal
from snownlp import seg
from snownlp.summary import textrank
if __name__ == '__main__':
t = normal.zh2hans(text)
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
rank = textrank.TextRank(doc)
rank.solve()
for index in rank.top_index(10):
print sents[index]
def getJson(fold, filename):
result = {}
try:
count = 0
cotent = u''
title = ''
time = ''
abstract = ''
path = fold + '/' + filename
# ========================================
# 读取文件的时间、标题、内容
# ========================================
for line in open(path, 'r'):
if (count == 0):
title = line
count += 1
# print (title)
continue
if (count == 1):
time = line
count += 1
# print (time)
continue
if (count > 1):
count += 1
cotent += line
# print (line)
# ========================================
# 生成摘要
# =======================================
t = normal.zh2hans(cotent)
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
rank = textrank.TextRank(doc)
rank.solve()
for index in rank.top_index(5):
abstract = abstract + sents[index] + ' '
keyword_rank = textrank.KeywordTextRank(doc)
keyword_rank.solve()
word0 = {}
word1 = {}
word2 = {}
word3 = {}
word4 = {}
wordcount = 0
for w in keyword_rank.top_index(5):
if wordcount == 0:
word0["word"] = w
word0["frequency"] = float(cotent.count(w)) / float(
len(cotent))
if wordcount == 1:
word1["word"] = w
word1["frequency"] = float(cotent.count(w)) / float(
len(cotent))
if wordcount == 2:
word2["word"] = w
word2["frequency"] = float(cotent.count(w)) / float(
len(cotent))
if wordcount == 3:
word3["word"] = w
word3["frequency"] = float(cotent.count(w)) / float(
len(cotent))
if wordcount == 4:
word4["word"] = w
word4["frequency"] = float(cotent.count(w)) / float(
len(cotent))
wordcount += 1
s = SnowNLP(cotent)
score = (s.sentiments - 0.5) * 2 # -1-1规范化
keywords = [word0, word1, word2, word3, word4]
result["code"] = 0
result["message"] = "sucess"
except IOError:
result["code"] = 1
result["message"] = "wrong format"
return result
result["tilte"] = title.strip()
result["time"] = time.strip()
result['abstract'] = abstract
result['sentiment'] = score
result["keywords"] = keywords
return result
6．用好你的相机
随着智能手机和平板电脑的普及，相机也变得无处不在，而且分享照片也越来越简单。MOOC的明星教授说，把45分钟的讲座变成10分钟一段的视频让他们被迫“升级课程”。不是每个老师都能通过这种方式吸引一批学生，但是他们可以参考这个经验，为课堂制作自己的视频，例如实地考察录像。让整个班都出去跑一趟可能不可行，但利用视频和照片，可以把考察点“带”到课室中来。利用智能手机耳机上配备的话筒，还可以为视频配上讲解，从而高效地用多个视频介绍完一个知识点。
将MOOC应用到传统课堂教学
随着大规模网络公开课的发展，教师可以考虑把在线教育的方法应用到自己的课堂教学中。MOOC的课程制作涉及比较复杂的技术，但使用这些课程几乎不费吹灰之力，而且成本也远远不及课程制作。没有加入edX或Coursera的大部分学校可以进行更多自创内容的尝试，就像自出版一样，这也是许多cMOOC的尝试。教师也可以向自己的目标努力。通过打开课堂，建立网络社区和制作教学视频，可以让更多的教师和学生享受到MOOC的投入带来的收益。
'''
from snownlp import normal
from snownlp import seg
from snownlp.summary import textrank
if __name__ == '__main__':
t = normal.zh2hans(text)
sents = normal.get_sentences(t)
doc = []
for sent in sents:
words = seg.seg(sent)
words = normal.filter_stop(words)
doc.append(words)
rank = textrank.TextRank(doc)
rank.solve()
for index in rank.top_index(5):
print(sents[index])
keyword_rank = textrank.KeywordTextRank(doc)
keyword_rank.solve()
for w in keyword_rank.top_index(5):
print(w)
