python期末大作业课程设计基于情感分析的新闻管理系统源码.zip

import datetime import json from operator import length_hint import time from django.conf import settings from django.shortcuts import render, redirect from django.contrib import messages from news.models import Comments, News, User, Category from django.http import JsonResponse from django.core.paginator import Paginator # Django内置分页功能模块 from django.core import serializers import os import jieba import numpy as np import tensorflow as tf import re from gensim.models import KeyedVectors # #加载预训练词向量 # cn_model = KeyedVectors.load_word2vec_format('/Users/lynn/Desktop/mysite/sgns.weibo.word.bz2', binary=False) # embedding_dim_c = cn_model['山东大学'].shape[0] # print("评论模型加载完毕") # n_model = KeyedVectors.load_word2vec_format('/Users/lynn/Desktop/mysite/sgns.sogou.word.bz2', binary=False) # embedding_dim_n = n_model['山东大学'].shape[0] # print("新闻模型加载完毕") # 文本处理（去除不必要的标签） def remove_tags(text): clean_text = re.compile(r'<[^>]+>') return clean_text.sub('', text) # 用于清理文本 def clean_text(x): x = str(x) x.replace('/>', ' /> ') for punct in "/-'": x = x.replace(punct, '') for punct in ' 一…了': x = x.replace(punct, '') for punct in '&': x = x.replace(punct, f' {punct} ') for punct in '?!.,"#$%\'()*+-/:;<=>@[\\]^_`{|}~丨「」|①④（）、｜:|' + '“”’': x = x.replace(punct, '') return x def data_process(train_reviews, max_len, cn_model, embedding_dim): print("评论：" + train_reviews[0]) # 进行分词和tokenize train_tokenize = [] for text in train_reviews: text = remove_tags(text) text = clean_text(text) cut = jieba.cut(text) cut_list = [i for i in cut] for i, word in enumerate(cut_list): try: cut_list[i] = cn_model.key_to_index[word] except KeyError: cut_list[i] = 0 train_tokenize.append(cut_list) # 反索引化功能 def reverse_tokens(tokens): text = '' for i in tokens: if i != 0: text = text + cn_model.index_to_key[i] else: text = text + '' return text reverse = reverse_tokens(train_tokenize[0]) print("反索引化的文本为：" + reverse) print("原始文本：" + train_reviews[0]) # 选取频率最高的50000词 num_words = 50000 embedding_matrix = np.zeros((num_words, embedding_dim)) for i in range(num_words): embedding_matrix[i, :] = cn_model[cn_model.index_to_key[i]] embedding_matrix = embedding_matrix.astype('float32') print(np.sum(cn_model[cn_model.index_to_key[333]] == embedding_matrix[333])) print(embedding_matrix.shape) # 填充和修剪 train_pad = tf.keras.preprocessing.sequence.pad_sequences(train_tokenize, maxlen=max_len, padding='post', truncating='post') # 超出50000个词向量的词用0代替 train_pad[train_pad >= num_words] = 0 return train_pad context={ 'user_num':User.objects.count(), 'comment_num':Comments.objects.count(), 'news_num':News.objects.count() } def index(request): # 用户登录 if request.method == "POST": username = request.POST.get('username', None) password = request.POST.get('password', None) if username and password: # 确保用户名和密码都不为空 username = username.strip() try: user = User.objects.get(user_name=username) if user.password == password: messages.add_message(request, messages.INFO, "登录成功！") request.session['is_login'] = True request.session['user_id'] = user.user_id request.session['user_name'] = user.user_name request.session['is_admin'] = user.is_admin request.session['login_time'] = str(datetime.datetime.now())[0:19] return redirect('/index') else: messages.add_message(request, messages.INFO, "密码不正确") except: return redirect('/index') # 新闻显示 physical=[] economy=[] entertain=[] education=[] technique=[] policy=[] news=News.objects.all() for new in news: if new.category_id==3: physical.append(new) elif new.category_id == 5: economy.append(new) elif new.category_id==4: entertain.append(new) elif new.category_id==7: education.append(new) elif new.category_id==8: technique.append(new) elif new.category_id==1: policy.append(new) news_show={'physical':physical[:5],'economy':economy[:18],'entertain':entertain[:18], 'education':education[:40], 'technique':technique[:18], 'policy':policy[:18]} return render(request, 'index.html', news_show) # 退出登录 def logout(request): request.session.flush() messages.add_message(request, messages.INFO, "已退出登录！") return redirect('/index') # 类别界面 def newsModule(request, category_id): newsList=[] news = News.objects.filter(category_id=category_id).all() n=0 for new in news: data=dict() data['id']=new.news_id data['title']=new.n_title data['time']=str(new.n_time)[0:19] data['content']=new.n_content[0:150]+'...' newsList.append(data) n+=1 if n==6: break return render(request, 'newsModule.html', {'newsList':newsList, "category_id":category_id}) def newsShow(request, news_id): news = News.objects.filter(news_id=news_id).all() for new in news: category=Category.objects.filter(category_id=new.category_id).all() for c in category: newsContent={ 'id':new.news_id, 'title':new.n_title, 'author':new.author, 'time':str(new.n_time)[0:19], 'content':new.n_content, 'category':new.category_id, 'category_name':c.category_name, 'img':new.img } comments=News.objects.get(news_id=news_id).comments_set.all() name_list=[] for comment in comments: name_list.append(User.objects.get(user_id=comment.user_id).user_name) comment_list=zip(name_list, comments) #提交用户评论 if request.method == "POST": user_id=request.session.get('user_id') if user_id is not None: content = request.POST.get('content', None) # 确定类别id id_list=[] for c in Comments.objects.all(): id_list.append(c.comment_id) comment_id=max_id(id_list) Comments.objects.create(comment_id=comment_id+1, comment_content=content, comment_time=str(datetime.datetime.now())[0:19], comment_senti=3, news_id=news_id, user_id=user_id) else: messages.add_message(request, messages.INFO, "请登录后发表评论") return redirect('/newsShow/'+str(news_id)) return render(request, 'newsShow.html', {'newsContent':newsContent, 'comment_list':comment_list}) def search(request): categories=Category.objects.all() return render(request, 'search.html',{'categories':categories}) # 新闻筛选 def searchData(request): category = request.GET.get("data[category]") year= request.GET.get("data[year]") senti= request.GET.get("data[senti]") news_list=News.objects.filter(n_senti=2).all() news_list=News.objects.all() if category=='' and year=='' and senti=='': news_list=Ne