def classify(self, results):
"""Classifies the results set by adding a "tag" attribute to each of the results.
The same set of results are returned, with additional statistics and tagging.
Each result gets one of the tags :), :( or :|
And a stats section is added.
@return an annotated array of results.
[{"tag": "pos",
"iso_language_code": "en",
"text": "@chucklelate im not that excited about google voice. although it seems neat, i dont see myself using it.",
"created_at": "Sat, 14 Mar 2009 00:00:03 +0000",
"profile_image_url": "http:\/\/s3.amazonaws.com\/twitter_production\/profile_images\/80373954\/IMG_0008_normal.JPG",
"to_user": "******",
"source": "<a href="http:\/\/twitter.com\/">web<\/a>",
"from_user": "******",
"from_user_id": 5160745,
"to_user_id": 409063,
"id": 1324759664},...],
"""
c = BayesianClassifier()
c.train(db.fetch_all_tweets())
for result in results:
tag = c.classify(result['text'])
result['tag'] = tag
return results
def classify(self, results):
"""Classifies the results set by adding a "tag" attribute to each of the results.
The same set of results are returned, with additional statistics and tagging.
Each result gets one of the tags :), :( or :|
And a stats section is added.
@return an annotated array of results.
[{"tag": "pos",
"iso_language_code": "en",
"text": "@chucklelate im not that excited about google voice. although it seems neat, i dont see myself using it.",
"created_at": "Sat, 14 Mar 2009 00:00:03 +0000",
"profile_image_url": "http:\/\/s3.amazonaws.com\/twitter_production\/profile_images\/80373954\/IMG_0008_normal.JPG",
"to_user": "******",
"source": "<a href="http:\/\/twitter.com\/">web<\/a>",
"from_user": "******",
"from_user_id": 5160745,
"to_user_id": 409063,
"id": 1324759664},...],
"""
c = BayesianClassifier()
c.train(db.fetch_all_tweets())
for result in results:
tag = c.classify(result['text'])
result['tag'] = tag
return results
def twitgraph(request):
bayesname = 'bayesdata'
srch = 'happy'
tag = 'n'
c = BayesianClassifier()
if not memcache.get('bayesdata'):
c.save() #laod form local file system
c.load() #load picke from memecache
'''q=request.get_all("q")'''
if (request.GET.has_key('q')):
srch = request.GET['q']
tag = c.classify(srch)
html = "<html><body>sentiment is %s </body></html>" % tag
#return HttpResponse(html)
gv = globvars
context = {'thispage': 'Sentiment', 'the_tweet': srch, 'sentmnt': tag}
context = dict(context, **gv.context)
return render_to_response('sentiment.html', context)
def classify(self, results):
c = BayesianClassifier()
c.train(db.fetch_all_status())
for result in results:
message = "%s" % result.get('message')
comments = result.get('comments')
tag = c.classify(message)
x = {str('sentiment'):str(tag)}
result.update(x)
if 'data' in comments:
for cm in comments.get('data'):
comment = "%s" % cm.get('message')
c_tag = c.classify(comment)
cm.update(sentiment=c_tag)
return results
def get(self):
id = self.request.get("id")
c = BayesianClassifier()
#access_token = self.request.cookies.get('at', '')
analyzer = StatusAnalyzer()
fetcher = StatusFetcher()
info_dict = fetcher.FetchInfo(id)
data = fetcher.FetchWall(id)
classified_results = analyzer.classify(data.get("data"))
template_values = {
'id': info_dict.id,
'feed': classified_results,
'picture': info_dict.picture,
'likes': info_dict.likes,
'name': info_dict.name,
'category': info_dict.category,
}
path = os.path.join(os.path.dirname(__file__),
'templates/analyze.html')
self.response.out.write(template.render(path, template_values))
def twitgraph(request):
bayesname='bayesdata'
srch = 'happy'
tag = 'n'
c = BayesianClassifier()
if not memcache.get('bayesdata'):
c.save()#laod form local file system
c.load()#load picke from memecache
'''q=request.get_all("q")'''
if(request.GET.has_key('q')):
srch = request.GET['q']
tag = c.classify(srch)
html = "<html><body>sentiment is %s </body></html>" % tag
#return HttpResponse(html)
gv = globvars
context = {
'thispage':'Sentiment',
'the_tweet':srch,
'sentmnt':tag
}
context = dict(context, **gv.context)
return render_to_response('sentiment.html', context)
