#!/usr/bin/env python import pickle import glob import textrank.ranker import textrank.tagger pickle_file = open("./textrank_tagger.pickle", "r+") tagger = pickle.load(pickle_file) pickle_file.close() #tagger = TextRankTagger(nltk.corpus.brown.tagged_sents()) ranker = textrank.ranker.TextRank(tagger) for fname in glob.glob("comments/*"): f = open(fname) comments = f.read() processed = ranker.preprocess(comments) #for phrase in ranker.extract_keywords(processed): # print "%s (%.4f)" % (str(phrase), phrase.score()) ranker.extract_sentences(processed, 10)
#!/usr/bin/env python import pickle import glob import textrank.ranker import textrank.tagger pickle_file = open("./textrank_tagger.pickle", "r+") tagger = pickle.load(pickle_file) pickle_file.close() # tagger = TextRankTagger(nltk.corpus.brown.tagged_sents()) ranker = textrank.ranker.TextRank(tagger) for fname in glob.glob("comments/*"): f = open(fname) comments = f.read() processed = ranker.preprocess(comments) # for phrase in ranker.extract_keywords(processed): # print "%s (%.4f)" % (str(phrase), phrase.score()) ranker.extract_sentences(processed, 10)
#!/usr/bin/env python import os import sys import pickle import glob import textrank.ranker import textrank.tagger pickle_file = open(os.path.dirname(__file__) + "/textrank_tagger.pickle") tagger = pickle.load(pickle_file) pickle_file.close() #tagger = TextRankTagger(nltk.corpus.brown.tagged_sents()) ranker = textrank.ranker.TextRank(tagger) comments = sys.stdin.read() ranker.extract_sentences(comments, 10)