The system structure of the network public opinion monitoring system

The system structure of the network public opinion monitoring system is the capture system and analysis system, and the capture system should be able to capture relevant information in real time.

Network public opinion is the mapping of social public opinion in the Internet space, and it is a direct reflection of social public opinion. Traditional social public opinion exists in the people, exists in the public's ideology and daily street discussions, the former is difficult to capture, the latter is fleeting, and the acquisition of public opinion can only be carried out through social investigation and secret visits, public opinion surveys, etc., the acquisition efficiency is low, the sample is small and easy to be biased, and the cost is huge. With the development of the Internet, the public often expresses their views in an information-based way, and network public opinion can be easily obtained by technical means such as automatic network capture, with high efficiency and information fidelity (no human processing), and full coverage.

Shanghai Mido's Sina Public Opinion Pass obtains, cleans, monitors, analyzes, and warns of network-wide data according to customer needs, and reduces the impact of human factors on objective data analysis results through data mining and analysis models, so as to ensure the timeliness, accuracy and comprehensiveness of public opinion data. The professional public opinion service team provides more personalized manual services according to the specific needs of customers, including: content sorting, important news push, briefing production, professional report customization and other public opinion services.

The network public opinion monitoring system is generally composed of an automatic collection subsystem (collection layer) and an analysis and browsing subsystem (analysis layer and presentation layer), taking the Lessee public opinion monitoring system as an example, the network public opinion monitoring system architecture includes three levels:

1. The collection layer, which includes element collection, keyword extraction, full-text indexing, automatic deduplication and differentiated storage and database, can collect microblogs, forums, blogs, post bars, news and comments, search engines, images and **, etc.

Second, the analysis layer can be changed to automatically classify, automatically cluster, automatically summarize, name recognition, positive and negative nature prediction and Chinese word segmentation operations for the collected data information to ensure the comprehensiveness of distraction.

3. The third layer is the presentation layer, and the data collected and analyzed by the system can be pushed to users in the form of negative public opinion, classified public opinion, latest public opinion, special tracking, public opinion briefing, classification evaluation, chart statistics and SMS notification.

The relationship is shown in the following diagram:

1. Public opinion collection system:

1. As long as the public opinion information related to "me" occurs on the Internet, it can be monitored at the first time and displayed in the most intuitive way, "all in one net, at a glance". Types of monitoring** include: news, forums, blogs, post bars, microblogs, newsletters, search engines, etc.

2. For key public opinion and negative information, timely warnings are given through mobile phone text messages and other means, and public opinion can be grasped at any time without the need for a special person on duty.

4. Automatically generate various statistical analysis reports and public opinion reports to help public opinion work. With the corresponding working mechanism, it can effectively improve the quality and efficiency of public opinion supervision and improve the level of public opinion response.

5. In addition to providing system-level 7*24-hour operation and maintenance services, it is also equipped with special public opinion analysts to assist in monitoring and manual early warning. The goal of system construction is to integrate Internet information channels and form a systematic and effective public opinion monitoring mechanism. Realize the operation of the system, monitor Internet information, Sina, Tencent and other major microblogs and microblogs, collect real-time data, monitor, analyze and retrieve the whole network, give early warning to sensitive information, prevent the spread of negative information, and make the most timely response and corresponding treatment suggestions for major events.

And search for hot issues and sensitive words and sentences in the recent period, so as to grasp the network public opinion and assist the leadership decision-making service. Major portals**, major newspapers, major large online forums, communities, post bars, blogs, microblogs. For example, Sina News, electronic newspapers of major newspapers, Tianya Forum, Sina Weibo, Tieba, etc.

2. Public opinion analysis system:

The analysis engine is a key component of the system. Its main function is to automatically analyze the data collected by the acquisition system. The main features of the analytics engine include:

Customize the analysis of public opinion level, automatic generation of hot spots, negative public opinion research and judgment, automatic classification, automatic generation of topics, ** calculation, automatic extraction of public opinion elements and keywords, automatic summary, automatic early warning, automatic generation of statistical charts and other functions. Similar to the system engine of Dorik public opinion data analysis station, it has built-in public opinion model, enterprise public opinion model and vertical monitoring model, which is constantly reconstructed and improved in accordance with the actual needs of customers in the public opinion industry for many years, and has good practical application effects. In the actual project, it can fully meet the needs of the enterprise without secondary development.

Or complete the vertical monitoring needs from top to bottom through a SaaS platform. For special application needs, the analysis engine also supports extension plug-ins for rapid secondary development.

3. Public opinion service platform:

It is mainly a platform for users to carry out daily public opinion management, which can receive public opinion information in a timely manner and carry out some routine public opinion management work.

Network public opinion monitoring generally needs to realize the analysis of the content of popular web pages, posts, and blog posts in important portals** forums and microblogs, and collect the number of reads and replies, and give early warnings in a timely manner.

A general system should include these major parts:

1. The content collection module generally includes the web crawler function and the public opinion interface function of the microblogging platform.

2. Semantic analysis and classification module, including compound keyword analysis function and natural language recognition function, to realize the analysis and classification of text content.

3. The data storage module stores the latest network information and analysis results in a structured manner. For information that needs to be stored for a long time, consider the dump function.

4. The hierarchical early warning and alarm module supports the analysis and classification of network content, according to the number of readings, the amount of reply, the amount of ** and other indicators, and realizes the alarm through the corresponding alarm channel.

6. Alarm and operation log management, realize the query and maintenance of historical alarms and operation logs.

7. System configuration and management module, realize management configuration, authority management configuration, public opinion monitoring list configuration, alarm threshold configuration, system function parameter configuration (such as docking parameters, function switches, etc.).

That's all I think of, and other netizens are welcome to add or correct them.