def suggestauthors(topic,num):
#read file and find the profs closest match to rawlabel and confirm/return actual label
authordict = getauthordict('authordict.csv')
ranking = []
similars = []
author = Analyze.idauthor(authordict,topic)
if author == '':
score=[]
stops=nltk.corpus.stopwords.words('english') #stopwords to weed out
stops = stops + ['we',',','.','(',')','using','new','propose','investigate']
stops = stops + ['-','show','infer','novel','method']
tokens1 = nltk.word_tokenize(topic)
pairs1 = nltk.bigrams(tokens1)
tokens1 = tokens1+[bg for bg in pairs1 if bg[0] not in stops and bg[1] not in stops]
for auth in authordict.keys():
keyw2 = authordict[auth]['Keywords']
tokens2 = list(filter(None,re.split(r',',keyw2)+re.split(r'[ ,]',keyw2)))
score = -sum(1 for token in tokens1 if token in tokens2)
heappush(ranking,[score,auth])
while len(similars)<num:
authscore = heappop(ranking)
similars.append([authscore[1],authordict[authscore[1]]['Keywords']])
else:
for auth in authordict:
score = Analyze.similarauthors(authordict[author],authordict[auth])
heappush(ranking,[score,auth])
while len(similars)<num:
authscore = heappop(ranking)
similars.append([authscore[1],authordict[authscore[1]]['Keywords']])
print(authordict[author]['Keywords'])
return similars
