28 lines
1.8 KiB
Markdown
28 lines
1.8 KiB
Markdown
## Configuration and Usage Instructions
|
|
|
|
After installing the application, please go to `application installation directory/config/config.yaml` to configure the remaining parameters.
|
|
|
|
After installing the application, please navigate to `application installation directory/config/frequency_words.txt` to configure the hot keywords you care about.
|
|
|
|
> After upgrading the application, the `config.yaml` needs to be updated as well.
|
|
> [Configuration Guide](https://github.com/sansan0/TrendRadar/blob/master/README-EN.md#%EF%B8%8F-configuration-guide)
|
|
> [Visual Configuration Editor](https://sansan0.github.io/TrendRadar)
|
|
|
|
Web Service: After entering the container, execute `python manage.py start_webserver` to start the service.
|
|
|
|
## Introduction
|
|
|
|
Simple public opinion monitoring and analysis - multi-platform hotspots aggregation + AI analysis tool based on MCP.
|
|
|
|
> Does not integrate AI by default, MCP service needs to be installed separately.
|
|
|
|
## Features
|
|
|
|
- **Aggregation of Hotspots Across the Network**: By default, monitors 11 major platforms such as Zhihu, Douyin, Bilibili, etc., and additional platforms can also be added manually;
|
|
- **Intelligent Push Strategy**: Supports three push modes: daily summary, current ranking, and incremental monitoring;
|
|
- **Precise Content Filtering**: Set personal keywords, only push relevant hotspots, filter irrelevant information;
|
|
- **Trend Analysis**: Real-time tracking of news popularity changes;
|
|
- **Personalized Hotspot Algorithm**: TrendRadar's proprietary weighted algorithm;
|
|
- **Multi-channel Real-time Push**: Supports WeCom (+ WeChat), Lark, DingTalk, Telegram, email, ntfy, and other push channels;
|
|
- **AI Intelligent Analysis**: Requires additional deployment of MCP service, refer to [documentation](https://github.com/sansan0/TrendRadar#-ai-%E6%99%BA%E8%83%BD%E5%88%86%E6%9E%90)
|