毕业设计 基于Python水利“舆情”构建知识图谱(Neo4j)应用源码+详细文档+全部数据资料（高分项目）.zip

#-*-coding:UTF-8-*- # import sys from pyltp import Segmentor from pyltp import Postagger from pyltp import Parser from pyltp import SentenceSplitter from pyltp import NamedEntityRecognizer import re import json seg = Segmentor() seg.load_with_lexicon('ltp_data_v3.4.0/cws.model', './ext_word') post = Postagger() post.load_with_lexicon('ltp_data_v3.4.0/pos.model', './ext_word_pos') recognizer = NamedEntityRecognizer() recognizer.load('ltp_data_v3.4.0/ner.model') parser = Parser() parser.load('ltp_data_v3.4.0/parser.model') def segmentor(sentence): """ 断词 :param sentence: 语句 :return: 词列表 """"" global seg words = seg.segment(sentence) words_list = list(words) return words_list def postagger(words): """ 获取词性 :param words: 词 :return: 词性 """ global post pos = post.postag(words) pos_list = list(pos) return pos_list def getdate(word): """ 获取月份 :param word: 词 :return: 月 """ if ("春" in word) or ("夏" in word) or ("秋" in word) or ("冬" in word): return word if "月" in word: _sentence = word.replace('十一月', '11月').replace('十二月', '12月') _sentence = _sentence.replace('一月', '1月').replace('二月', '2月').replace('三月', '3月'). \ replace('四月', '4月').replace('五月', '5月') _sentence = _sentence.replace('六月', '6月').replace('七月', '7月').replace('八月', '8月'). \ replace('九月', '9月').replace('十月', '10月') else: _sentence = word m = re.search(r"(\d+)月", _sentence) if m is not None: _m = int(m.group(0).split("月")[0]) if (_m < 13) and (_m > 0): return _m return None def hasSBV(arcs): sbv = False vob = False hed = False for nn in arcs: if "SBV" in nn.relation: sbv = True if "VOB" in nn.relation: vob = True if "HED" in nn.relation: hed = True if (hed and sbv) or (vob and sbv): return True return False def chs_replace(sentence): """ :param sentence: :return: """ _s = sentence.replace("零", "0") _s = _s.replace("一", "1").replace("二", "2").replace("三", "3") _s = _s.replace("四", "4").replace("五", "5").replace("六", "6") _s = _s.replace("七", "7").replace("八", "8").replace("九", "9") _n_idx = 0 while True: _idx = _s.find("十", _n_idx) if _idx < 0: break if _idx == _n_idx: break if _idx > 0: if _s[_idx-1:_idx].isdigit(): if _s[_idx+3:_idx+4].isdigit(): _s = _s[:_idx] + _s[_idx+3:] else: _s = _s[:_idx] + "0" + _s[_idx+3:] else: if _s[_idx+3:_idx+4].isdigit(): _s = _s[:_idx] + "1" + _s[_idx+3:] _n_idx = _idx return _s year = None month = None events = {} line_cnt = 0 sentence_cnt = 0 """获取文本文件""" f = open(sys.argv[1]) while True: """读一段文本""" sentence = f.readline() if (sentence is "") or (len(sentence) == 0): """ Eof """ break """替代中文引号""" sentence = sentence.replace("‘", "").replace("’", "").replace("”", "").replace("“", "") line_cnt += 1 """断句""" sents = SentenceSplitter.split(sentence) for ss in sents: s = chs_replace(ss) if re.search(r"(\d+)年(\d+)月(\d+)日", s) is None: continue """语句处理""" sentence_cnt += 1 words = segmentor(s) pos = postagger(words) netags = list(recognizer.recognize(words, pos)) arcs = list(parser.parse(words, pos)) _i = 0 for _p in pos: # print _p, " : ", words[_i] if "nt" in _p: if "年" in words[_i]: _y = words[_i].split("年")[0] if str(_y).isdigit(): _y = int(_y) if (_y > 1000) and (_y < 2024): year = _y else: _m = getdate(words[_i]) if _m is not None: month = _m """Next""" _i += 1 if hasSBV(arcs) and (year is not None): if year not in events: events[year] = {} if month not in events[year]: events[year][month] = [] events[year][month].append(ss) """ _i = 0 for nn in arcs: print "%d:%s " % (nn.head, nn.relation), if nn.head > 0: print words[_i], " <---> ", words[nn.head-1] else: print words[_i] _i += 1 if "SBV" is nn.relation: if nn not in events[year][month]: events[year][month].append("%d:%s" % (nn.head, nn.relation)) """ # print(">>> %d, %d" % (line_cnt, sentence_cnt)) out_info = {} for _date in sorted(events, key=lambda x: x): print("<<< %s >>>" % str(_date)) if _date not in out_info: out_info[_date] = [] for _m in sorted(events[_date], key=lambda x: x): for _evt in events[_date][_m]: print("\t%-22s" % _evt) out_info[_date].append(_evt) parser.release() post.release() seg.release() with open("out.json", "w") as f: f.write(json.dumps(out_info)) f.close() print("Done.")