💖💖作者：计算机编程小咖 💙💙个人简介：曾长期从事计算机专业培训教学，本人也热爱上课教学，语言擅长Java、微信小程序、Python、Golang、安卓Android等，开发项目包括大数据、深度学习、网站、小程序、安卓、算法。平常会做一些项目定制化开发、代码讲解、答辩教学、文档编写、也懂一些降重方面的技巧。平常喜欢分享一些自己开发中遇到的问题的解决办法，也喜欢交流技术，大家有技术代码这一块的问题可以问我！ 💛💛想说的话：感谢大家的关注与支持！ 💜💜 网站实战项目 安卓/小程序实战项目 大数据实战项目 深度学习实战项目
@TOC
基于大数据的社交媒体舆情数据可视化分析系统介绍
《基于大数据的社交媒体舆情数据可视化分析系统》是一个集数据采集、处理、分析与可视化于一体的综合性舆情监测平台，该系统充分运用Hadoop分布式存储和Spark大数据处理技术，构建了高效的海量社交媒体数据处理架构，通过HDFS实现数据的分布式存储，利用Spark SQL进行快速数据查询与分析，结合Pandas和NumPy进行深度数据挖掘和统计计算。系统采用前后端分离的技术架构，后端基于Django或Spring Boot框架构建RESTful API服务，前端运用Vue.js结合ElementUI组件库打造现代化的用户界面，通过Echarts图表库实现丰富的数据可视化效果，整体界面采用HTML、CSS、JavaScript和jQuery技术确保良好的用户体验。系统核心功能模块包括系统首页展示平台概况、我的信息进行用户个人管理、系统管理实现平台运维，更重要的是提供大屏可视化实时展示舆情态势、数据分析模块深度挖掘数据价值、舆情总览分析全面掌握舆情动态、情感趋势分析追踪公众情感变化、内容画像分析识别热点话题特征，以及用户与互动分析洞察用户行为模式，整个系统通过MySQL数据库存储结构化数据，为企业、政府部门和研究机构提供全面、准确、及时的社交媒体舆情监测和分析服务，助力决策者快速掌握网络舆论走向和公众情感态度。
基于大数据的社交媒体舆情数据可视化分析系统演示视频
基于大数据的社交媒体舆情数据可视化分析系统演示图片
基于大数据的社交媒体舆情数据可视化分析系统代码展示
spark = SparkSession.builder.appName("SocialMediaSentimentAnalysis").config("spark.sql.adaptive.enabled", "true").config("spark.sql.adaptive.coalescePartitions.enabled", "true").getOrCreate()
def analyze_sentiment_overview(request):
social_data_df = spark.read.option("header", "true").csv("hdfs://localhost:9000/social_media_data/*.csv")
social_data_df.createOrReplaceTempView("social_media_posts")
total_posts = spark.sql("SELECT COUNT(*) as total FROM social_media_posts").collect()[0]['total']
positive_posts = spark.sql("SELECT COUNT(*) as count FROM social_media_posts WHERE sentiment_score > 0.6").collect()[0]['count']
negative_posts = spark.sql("SELECT COUNT(*) as count FROM social_media_posts WHERE sentiment_score < 0.4").collect()[0]['count']
neutral_posts = total_posts - positive_posts - negative_posts
platform_stats = spark.sql("SELECT platform, COUNT(*) as count, AVG(sentiment_score) as avg_sentiment FROM social_media_posts GROUP BY platform").collect()
hot_keywords = spark.sql("SELECT keyword, COUNT(*) as frequency FROM social_media_posts LATERAL VIEW explode(split(content, ' ')) t AS keyword WHERE length(keyword) > 2 GROUP BY keyword ORDER BY frequency DESC LIMIT 20").collect()
sentiment_distribution = spark.sql("SELECT CASE WHEN sentiment_score > 0.6 THEN 'positive' WHEN sentiment_score < 0.4 THEN 'negative' ELSE 'neutral' END as sentiment_type, COUNT(*) as count FROM social_media_posts GROUP BY sentiment_type").collect()
daily_posts = spark.sql("SELECT date_format(post_time, 'yyyy-MM-dd') as date, COUNT(*) as count FROM social_media_posts WHERE post_time >= date_sub(current_date(), 30) GROUP BY date ORDER BY date").collect()
engagement_stats = spark.sql("SELECT AVG(likes_count) as avg_likes, AVG(comments_count) as avg_comments, AVG(shares_count) as avg_shares FROM social_media_posts").collect()[0]
high_influence_posts = spark.sql("SELECT content, sentiment_score, likes_count + comments_count + shares_count as total_engagement FROM social_media_posts WHERE likes_count + comments_count + shares_count > 1000 ORDER BY total_engagement DESC LIMIT 10").collect()
sentiment_by_time = spark.sql("SELECT hour(post_time) as hour, AVG(sentiment_score) as avg_sentiment FROM social_media_posts GROUP BY hour ORDER BY hour").collect()
topic_sentiment = spark.sql("SELECT topic_category, COUNT(*) as post_count, AVG(sentiment_score) as avg_sentiment FROM social_media_posts WHERE topic_category IS NOT NULL GROUP BY topic_category ORDER BY post_count DESC").collect()
user_activity = spark.sql("SELECT user_id, COUNT(*) as post_count, AVG(sentiment_score) as user_sentiment FROM social_media_posts GROUP BY user_id HAVING post_count >= 5 ORDER BY post_count DESC LIMIT 50").collect()
regional_sentiment = spark.sql("SELECT region, COUNT(*) as count, AVG(sentiment_score) as avg_sentiment FROM social_media_posts WHERE region IS NOT NULL GROUP BY region ORDER BY count DESC").collect()
content_length_sentiment = spark.sql("SELECT CASE WHEN length(content) < 50 THEN 'short' WHEN length(content) < 200 THEN 'medium' ELSE 'long' END as content_type, AVG(sentiment_score) as avg_sentiment FROM social_media_posts GROUP BY content_type").collect()
overview_data = {'total_posts': total_posts, 'positive_ratio': round(positive_posts/total_posts*100, 2), 'negative_ratio': round(negative_posts/total_posts*100, 2), 'neutral_ratio': round(neutral_posts/total_posts*100, 2), 'platform_stats': [{'platform': row['platform'], 'count': row['count'], 'sentiment': round(row['avg_sentiment'], 3)} for row in platform_stats], 'hot_keywords': [{'keyword': row['keyword'], 'frequency': row['frequency']} for row in hot_keywords], 'daily_trend': [{'date': row['date'], 'count': row['count']} for row in daily_posts], 'engagement_avg': {'likes': round(engagement_stats['avg_likes'], 1), 'comments': round(engagement_stats['avg_comments'], 1), 'shares': round(engagement_stats['avg_shares'], 1)}, 'high_influence': [{'content': row['content'][:100], 'sentiment': round(row['sentiment_score'], 3), 'engagement': row['total_engagement']} for row in high_influence_posts]}
return JsonResponse(overview_data)
def analyze_sentiment_trend(request):
time_range = request.GET.get('range', '30')
social_data_df = spark.read.option("header", "true").csv("hdfs://localhost:9000/social_media_data/*.csv")
social_data_df.createOrReplaceTempView("social_trend_data")
daily_sentiment = spark.sql(f"SELECT date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as avg_sentiment, COUNT(*) as post_count FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY date ORDER BY date").collect()
hourly_sentiment = spark.sql(f"SELECT hour(post_time) as hour, AVG(sentiment_score) as avg_sentiment, COUNT(*) as post_count FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY hour ORDER BY hour").collect()
sentiment_volatility = spark.sql(f"SELECT date_format(post_time, 'yyyy-MM-dd') as date, stddev(sentiment_score) as sentiment_std FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY date ORDER BY date").collect()
platform_trend = spark.sql(f"SELECT platform, date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as avg_sentiment FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY platform, date ORDER BY platform, date").collect()
emotion_distribution = spark.sql(f"SELECT date_format(post_time, 'yyyy-MM-dd') as date, SUM(CASE WHEN sentiment_score > 0.6 THEN 1 ELSE 0 END) as positive, SUM(CASE WHEN sentiment_score < 0.4 THEN 1 ELSE 0 END) as negative, SUM(CASE WHEN sentiment_score >= 0.4 AND sentiment_score <= 0.6 THEN 1 ELSE 0 END) as neutral FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY date ORDER BY date").collect()
peak_sentiment_days = spark.sql(f"SELECT date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as avg_sentiment, COUNT(*) as post_count FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY date HAVING post_count > 100 ORDER BY avg_sentiment DESC LIMIT 5").collect()
low_sentiment_days = spark.sql(f"SELECT date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as avg_sentiment, COUNT(*) as post_count FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY date HAVING post_count > 100 ORDER BY avg_sentiment ASC LIMIT 5").collect()
weekly_comparison = spark.sql(f"SELECT weekofyear(post_time) as week, AVG(sentiment_score) as avg_sentiment, COUNT(*) as post_count FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY week ORDER BY week").collect()
sentiment_momentum = []
for i in range(1, len(daily_sentiment)):
current_sentiment = daily_sentiment[i]['avg_sentiment']
previous_sentiment = daily_sentiment[i-1]['avg_sentiment']
momentum = round((current_sentiment - previous_sentiment) / previous_sentiment * 100, 2) if previous_sentiment != 0 else 0
sentiment_momentum.append({'date': daily_sentiment[i]['date'], 'momentum': momentum})
keyword_sentiment_trend = spark.sql(f"SELECT keyword, date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as avg_sentiment FROM social_trend_data LATERAL VIEW explode(split(content, ' ')) t AS keyword WHERE length(keyword) > 3 AND post_time >= date_sub(current_date(), {time_range}) GROUP BY keyword, date HAVING COUNT(*) >= 10 ORDER BY keyword, date").collect()
user_sentiment_evolution = spark.sql(f"SELECT user_id, date_format(post_time, 'yyyy-MM-dd') as date, AVG(sentiment_score) as user_sentiment FROM social_trend_data WHERE post_time >= date_sub(current_date(), {time_range}) GROUP BY user_id, date HAVING COUNT(*) >= 3 ORDER BY user_id, date").collect()
trend_data = {'daily_sentiment': [{'date': row['date'], 'sentiment': round(row['avg_sentiment'], 3), 'posts': row['post_count']} for row in daily_sentiment], 'hourly_pattern': [{'hour': row['hour'], 'sentiment': round(row['avg_sentiment'], 3)} for row in hourly_sentiment], 'volatility': [{'date': row['date'], 'std': round(row['sentiment_std'], 3)} for row in sentiment_volatility], 'platform_trends': {}, 'emotion_timeline': [{'date': row['date'], 'positive': row['positive'], 'negative': row['negative'], 'neutral': row['neutral']} for row in emotion_distribution], 'peak_days': [{'date': row['date'], 'sentiment': round(row['avg_sentiment'], 3)} for row in peak_sentiment_days], 'momentum': sentiment_momentum}
for row in platform_trend:
platform = row['platform']
if platform not in trend_data['platform_trends']:
trend_data['platform_trends'][platform] = []
trend_data['platform_trends'][platform].append({'date': row['date'], 'sentiment': round(row['avg_sentiment'], 3)})
return JsonResponse(trend_data)
def generate_dashboard_visualization(request):
social_data_df = spark.read.option("header", "true").csv("hdfs://localhost:9000/social_media_data/*.csv")
social_data_df.createOrReplaceTempView("dashboard_data")
real_time_metrics = spark.sql("SELECT COUNT(*) as total_posts, AVG(sentiment_score) as avg_sentiment, SUM(likes_count + comments_count + shares_count) as total_engagement FROM dashboard_data WHERE post_time >= date_sub(current_timestamp(), interval 1 hour)").collect()[0]
sentiment_gauge = spark.sql("SELECT CASE WHEN sentiment_score > 0.6 THEN 'positive' WHEN sentiment_score < 0.4 THEN 'negative' ELSE 'neutral' END as sentiment, COUNT(*) as count FROM dashboard_data WHERE post_time >= date_sub(current_date(), 1) GROUP BY sentiment").collect()
geographic_distribution = spark.sql("SELECT region, COUNT(*) as post_count, AVG(sentiment_score) as avg_sentiment FROM dashboard_data WHERE region IS NOT NULL AND post_time >= date_sub(current_date(), 7) GROUP BY region ORDER BY post_count DESC LIMIT 20").collect()
trending_topics = spark.sql("SELECT topic_category, COUNT(*) as frequency, AVG(sentiment_score) as topic_sentiment FROM dashboard_data WHERE topic_category IS NOT NULL AND post_time >= date_sub(current_date(), 1) GROUP BY topic_category ORDER BY frequency DESC LIMIT 10").collect()
platform_performance = spark.sql("SELECT platform, COUNT(*) as posts, AVG(likes_count) as avg_likes, AVG(sentiment_score) as platform_sentiment FROM dashboard_data WHERE post_time >= date_sub(current_date(), 7) GROUP BY platform ORDER BY posts DESC").collect()
influence_ranking = spark.sql("SELECT user_id, COUNT(*) as post_count, SUM(likes_count + comments_count + shares_count) as total_influence, AVG(sentiment_score) as user_sentiment FROM dashboard_data WHERE post_time >= date_sub(current_date(), 7) GROUP BY user_id ORDER BY total_influence DESC LIMIT 15").collect()
content_analysis = spark.sql("SELECT CASE WHEN length(content) < 50 THEN 'short' WHEN length(content) < 200 THEN 'medium' ELSE 'long' END as content_length, COUNT(*) as count, AVG(sentiment_score) as avg_sentiment, AVG(likes_count + comments_count + shares_count) as avg_engagement FROM dashboard_data GROUP BY content_length").collect()
time_distribution = spark.sql("SELECT hour(post_time) as hour, COUNT(*) as post_count, AVG(sentiment_score) as hourly_sentiment FROM dashboard_data WHERE post_time >= date_sub(current_date(), 7) GROUP BY hour ORDER BY hour").collect()
sentiment_correlation = spark.sql("SELECT likes_count, comments_count, shares_count, sentiment_score FROM dashboard_data WHERE post_time >= date_sub(current_date(), 30) AND likes_count > 0").collect()
correlation_data = []
if sentiment_correlation:
likes_sentiment_corr = spark.sql("SELECT corr(likes_count, sentiment_score) as correlation FROM dashboard_data WHERE post_time >= date_sub(current_date(), 30)").collect()[0]['correlation']
comments_sentiment_corr = spark.sql("SELECT corr(comments_count, sentiment_score) as correlation FROM dashboard_data WHERE post_time >= date_sub(current_date(), 30)").collect()[0]['correlation']
correlation_data = [{'metric': 'likes', 'correlation': round(likes_sentiment_corr, 3)}, {'metric': 'comments', 'correlation': round(comments_sentiment_corr, 3)}]
alert_conditions = spark.sql("SELECT COUNT(*) as negative_spike FROM dashboard_data WHERE sentiment_score < 0.3 AND post_time >= date_sub(current_timestamp(), interval 2 hour)").collect()[0]['negative_spike']
high_engagement_posts = spark.sql("SELECT content, sentiment_score, likes_count + comments_count + shares_count as engagement FROM dashboard_data WHERE post_time >= date_sub(current_date(), 1) ORDER BY engagement DESC LIMIT 8").collect()
dashboard_data = {'real_time': {'posts': real_time_metrics['total_posts'], 'sentiment': round(real_time_metrics['avg_sentiment'], 3), 'engagement': real_time_metrics['total_engagement']}, 'sentiment_distribution': [{'type': row['sentiment'], 'count': row['count']} for row in sentiment_gauge], 'geographic_map': [{'region': row['region'], 'posts': row['post_count'], 'sentiment': round(row['avg_sentiment'], 3)} for row in geographic_distribution], 'trending_topics': [{'topic': row['topic_category'], 'frequency': row['frequency'], 'sentiment': round(row['topic_sentiment'], 3)} for row in trending_topics], 'platform_stats': [{'platform': row['platform'], 'posts': row['posts'], 'avg_likes': round(row['avg_likes'], 1), 'sentiment': round(row['platform_sentiment'], 3)} for row in platform_performance], 'top_influencers': [{'user': row['user_id'], 'influence': row['total_influence'], 'sentiment': round(row['user_sentiment'], 3)} for row in influence_ranking], 'content_analysis': [{'length': row['content_length'], 'count': row['count'], 'sentiment': round(row['avg_sentiment'], 3), 'engagement': round(row['avg_engagement'], 1)} for row in content_analysis], 'hourly_activity': [{'hour': row['hour'], 'posts': row['post_count'], 'sentiment': round(row['hourly_sentiment'], 3)} for row in time_distribution], 'alerts': {'negative_spike': alert_conditions > 50}, 'correlations': correlation_data, 'hot_posts': [{'content': row['content'][:80], 'sentiment': round(row['sentiment_score'], 3), 'engagement': row['engagement']} for row in high_engagement_posts]}
return JsonResponse(dashboard_data)
基于大数据的社交媒体舆情数据可视化分析系统文档展示
💖💖作者：计算机编程小咖 💙💙个人简介：曾长期从事计算机专业培训教学，本人也热爱上课教学，语言擅长Java、微信小程序、Python、Golang、安卓Android等，开发项目包括大数据、深度学习、网站、小程序、安卓、算法。平常会做一些项目定制化开发、代码讲解、答辩教学、文档编写、也懂一些降重方面的技巧。平常喜欢分享一些自己开发中遇到的问题的解决办法，也喜欢交流技术，大家有技术代码这一块的问题可以问我！ 💛💛想说的话：感谢大家的关注与支持！ 💜💜 网站实战项目 安卓/小程序实战项目 大数据实战项目 深度学习实战项目
