def analyzeNewTweets(query, N):
#normalize input, query should be short, N should not be larger
#than 500. ?not implemented?
#collect no more than N data
tweets = searchAndExtractTweets(query, N)
print "extracted " + str(len(tweets)) + " tweets"
# text summarization
summerizer = NaiveSummarizer()
textdoc = ""
for tweet in tweets:
textdoc += (tweet["text"] + ". ")
summ = summerizer.summarize(textdoc, 4)
print "summarize: " + summ
# sentiment analysis
payloaddata = json.dumps({'language': 'en', 'data': tweets})
req = urllib2.Request(url="http://www.sentiment140.com/api/[email protected]", data = payloaddata)
res = urllib2.urlopen(req)
# results into json
res = res.read().decode("UTF-8", 'ignore')
resdict = json.loads(res)
negList = []
neuList = []
posList = []
for sent in resdict["data"]:
if sent["polarity"] == 0:
negList.append(sent["text"])
if sent["polarity"] == 4:
posList.append(sent["text"])
if sent["polarity"] == 2:
neuList.append(sent["text"])
return {"positive": posList, "neutral": neuList, "negative": negList, "summary": summ}
def analyzeNewFBPosts(topic, number, token):
payloaddata = retriveData(topic, number, token)
# summary
summerizer = NaiveSummarizer()
textdoc = ""
for data in payloaddata:
textdoc += (data + ". ")
summ = summerizer.summarize(textdoc, 4)
print "summarize: " + summ
# result into json
posList = []
neuList = []
negList = []
index = 0
for post in payloaddata:
res = getSentiment(post)
if res is not "none":
if res == "positive":
posList.append(post)
if res == "negative":
negList.append(post)
# if res == "neutral":
# neuList.append(post)
index += 1
print str(index) + "/" + str(len(payloaddata))
# print json.dumps(results, indent = 4)
return {"positive": posList, "negative": negList, "neutral": neuList, "summary": summ}
# analyzeNewPosts('AI', 'CAACEdEose0cBAJPsfkUp71nxbZAFfCjgE1VPZAN3h5vE68T5PD9Ju00d8eedNqKaooZAVCU7AbYJYZAWNJ1sWfXO2WOMFHu4qxrS75GIUattFwZB8x0g4yoqzzZBxZCSqcyQnRUz5IrB3B9gKyvQfSf1UHKZBGNgcNFikjg3OhNK5pjYdM1ZALmu0Y6yuc5iCXXObtHEJkNaXfwZDZD')