def run(self):
self.ELMO = word_emb_elmo.WordEmbeddings(self.model_file, cuda_device=self.gpu_id)
self.SIF = sent_emb_sif.SentEmbeddings(self.ELMO, lamda=1.0)
if self.cut_dict == True:
self.zh_model = thulac.thulac(model_path=r'./auxiliary_data/thulac.models/', seg_only=self.seg_only)
else:
self.zh_model = thulac.thulac(model_path=r'./auxiliary_data/thulac.models/',user_dict=self.user_dict_file, seg_only=self.seg_only)
while self.stop_sign.value == 0:
if self.recv_queue.empty() == False:
try:
data = self.recv_queue.get(True, 1)
except Exception as e:
continue
docid = data[0]
text = data[1]
if len(text) > 4000:
text = text[0:4000]
# [title, content] = text.split('\t')
self.logger.info("worker_process[%d] %s, len:%d" %(self.worker_id, docid, len(text)))
keywords = extract_keyword(text, self.SIF, self.zh_model, self.elmo_layers_weight, plus=self.plus,
topk=20, kwdict=self.user_dict, kw_info=self.kw_info, cut_dict=self.cut_dict, seg_only=self.seg_only)
self.logger.info("worker_succ[%d] %s" %(self.worker_id, docid))
self.logger.info("worker_succ[%d] %s %s" %(self.worker_id, docid, keywords))
#self.push_queue.put([docid, title_kw, content_kw])
self.push_queue.put([docid, keywords])
self.logger.info("stop worker[%d]" %(self.worker_id))
def __init__(self):
# path = os.path.dirname(os.path.realpath('__file__'))
self.vncorenlp = VnCoreNLP(
"auxiliary_data/VnCoreNLP-master/VnCoreNLP-1.1.1.jar",
annotators="wseg, pos",
max_heap_size='-Xmx500m')
self.phoBERT = word_emb_phoBert.WordEmbeddings()
self.SIF = sent_emb_sif.SentEmbeddings(self.phoBERT,
lamda=1.0,
embeddings_type='bert')
def extract_article_kw(article_file, output_file):
docids, texts = load_articles(article_file)
#docids, texts = load_tencent_articles(article_file)
wfp = open(output_file, "w", encoding="utf-8")
user_dict_file = r'./auxiliary_data/keyword_vocab_final'
kw_info_file = r'/search/odin/liruihong/keyword-project/config_data/ret_item_info'
model_file = r'./auxiliary_data/zhs.model/'
gpu_id = 0
kwdict = load_user_dict(user_dict_file)
kw_info = load_kw_info(kw_info_file, encoding="gbk")
user_dict = load_user_dict(
"/search/odin/liruihong/keyword-project/keywords_embrank/config_data/articles_7d_vocab.txt"
)
seg_only = True
# ELMO = word_emb_elmo.WordEmbeddings(model_file, cuda_device=gpu_id)
# SIF = sent_emb_sif.SentEmbeddings(ELMO, lamda=1.0)
w2v = word2vec_emb.Word2VecEmbeddings(
"/search/odin/liruihong/keyword-project/keywords_embrank/config_data/articles_7d_word2vec.txt"
)
SIF = sent_emb_sif.SentEmbeddings(w2v, lamda=1.0, embeddings_type="glove")
# zh_model = thulac.thulac(model_path=r'./auxiliary_data/thulac.models/', seg_only=seg_only)
elmo_layers_weight = [1.0, 0.0, 0.0]
plus = True
kw_info = kw_info
checkfp = open(
"/search/odin/liruihong/keyword-project/SIFRank_zh/check_w2vsif_model.bin",
"wb")
check_data = {}
for idx, text in enumerate(texts):
#if idx >= 200:
# break
docid = docids[idx]
logging.info("%s" % (docid))
st = time.time()
keywords = extract_keyword(text,
SIF,
zh_model,
elmo_layers_weight,
plus=plus,
topk=20,
kwdict=kwdict,
kw_info=kw_info,
cut_dict=False,
seg_only=seg_only,
user_dict=user_dict)
if isinstance(keywords, tuple):
check_item = keywords[1]
check_data[docid] = check_item
keywords = keywords[0]
ed = time.time()
cost = int((ed - st) * 1000)
print("%s time_cost:%d ms" % (docid, cost))
#keywords = [(kw,score) for kw,score in keywords if kw in user_dict]
writer_keywords = " ".join(
["%s:%f" % (kw, score) for kw, score in keywords])
wfp.write("%s\t%s\n" % (docid, writer_keywords))
pickle.dump(check_data, checkfp)
checkfp.close()
wfp.close()
#! /usr/bin/env python
# -*- coding: utf-8 -*-
# __author__ = "Sponge_sy"
# Date: 2020/2/21
from embeddings import sent_emb_sif, word_emb_elmo
from model.method import SIFRank, SIFRank_plus
import thulac
import jieba.analyse
#download from https://github.com/HIT-SCIR/ELMoForManyLangs
model_file = r'../auxiliary_data/zhs.model/'
ELMO = word_emb_elmo.WordEmbeddings(model_file)
SIF = sent_emb_sif.SentEmbeddings(ELMO, lamda=1.0)
#download from http://thulac.thunlp.org/
zh_model = thulac.thulac(model_path=r'../auxiliary_data/thulac.models/',user_dict=r'../auxiliary_data/user_dict.txt')
elmo_layers_weight = [0.0, 1.0, 0.0]
text = "计算机科学与技术（Computer Science and Technology）是国家一级学科，下设信息安全、软件工程、计算机软件与理论、计算机系统结构、计算机应用技术、计算机技术等专业。 [1]主修大数据技术导论、数据采集与处理实践（Python）、Web前/后端开发、统计与数据分析、机器学习、高级数据库系统、数据可视化、云计算技术、人工智能、自然语言处理、媒体大数据案例分析、网络空间安全、计算机网络、数据结构、软件工程、操作系统等课程，以及大数据方向系列实验，并完成程序设计、数据分析、机器学习、数据可视化、大数据综合应用实践、专业实训和毕业设计等多种实践环节。"
keyphrases = SIFRank(text, SIF, zh_model, N=15,elmo_layers_weight=elmo_layers_weight)
keyphrases_ = SIFRank_plus(text, SIF, zh_model, N=15, elmo_layers_weight=elmo_layers_weight)
print("------------------------------------------")
print("原文:"+text)
print("------------------------------------------")
print("SIFRank_zh结果:")
print(keyphrases)
print("SIFRank+_zh结果:")
print(keyphrases_)
print("------------------------------------------")
print("jieba分词TFIDF算法结果:")
elif (database == "Duc2001"):
data, labels = fileIO.get_duc2001_data()
lamda = 1.0
elmo_layers_weight = [1.0, 0.0, 0.0]
else:
data, labels = fileIO.get_semeval2017_data()
lamda = 0.6
elmo_layers_weight = [1.0, 0.0, 0.0]
#download from https://allennlp.org/elmo
options_file = "../auxiliary_data/elmo_2x4096_512_2048cnn_2xhighway_options.json"
weight_file = "../auxiliary_data/elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5"
porter = nltk.PorterStemmer() #please download nltk
ELMO = word_emb_elmo.WordEmbeddings(options_file, weight_file, cuda_device=0)
SIF = sent_emb_sif.SentEmbeddings(ELMO, lamda=lamda, database=database)
en_model = StanfordCoreNLP(
r'E:\Python_Files\stanford-corenlp-full-2018-02-27',
quiet=True) #download from https://stanfordnlp.github.io/CoreNLP/
try:
for key, data in data.items():
lables = labels[key]
lables_stemed = []
for lable in lables:
tokens = lable.split()
lables_stemed.append(' '.join(porter.stem(t) for t in tokens))
print(key)
import sys
from flask import Flask, jsonify, request, render_template
from werkzeug.utils import secure_filename
import json
import os
from vncorenlp import VnCoreNLP
from embeddings import sent_emb_sif, word_emb_phoBert
from model.method import SIFRank, SIFRank_plus
app = Flask(__name__, static_url_path="", static_folder='./static/')
vncorenlp = VnCoreNLP("auxiliary_data/VnCoreNLP-master/VnCoreNLP-1.1.1.jar",
annotators="wseg, pos",
max_heap_size='-Xmx500m')
phoBERT = word_emb_phoBert.WordEmbeddings()
SIF = sent_emb_sif.SentEmbeddings(phoBERT, lamda=1.0, embeddings_type='bert')
# Xử lý điều hướng câu truy vấn từ client gửi đến
@app.route('/extract', methods=['GET'])
def extract():
return render_template('extract.html')
@app.route('/result', methods=['POST'])
def result():
text = request.form.get('query')
highlighted, phrases = extract_keyphrase(text)
return render_template('result.html',
highlighted=highlighted,
phrases=phrases)
data = request.json
query = data.get("text", "")
top_n = int(data.get("n", 15))
keywords = SIFRank_plus(query,
SIF,
en_model,
N=top_n,
elmo_layers_weight=elmo_layers_weight)
return jsonify(keywords)
if __name__ == '__main__':
data_dir = "data/"
options_file = f"{data_dir}elmo_2x4096_512_2048cnn_2xhighway_options.json"
weight_file = f"{data_dir}elmo_2x4096_512_2048cnn_2xhighway_weights.hdf5"
cuda_device = 0
porter = nltk.PorterStemmer()
ELMO = word_emb_elmo.WordEmbeddings(options_file=options_file,
weight_file=weight_file,
cuda_device=cuda_device)
SIF = sent_emb_sif.SentEmbeddings(word_embeddor=ELMO,
data_dir=data_dir,
lamda=1.0)
en_model = stanza.Pipeline(lang='en', processors={}, use_gpu=True)
elmo_layers_weight = [0.0, 1.0, 0.0]
serve(app, host="0.0.0.0", port=5000)
