Generate HN Community Insights & Sentiment with AI Clustering
Uncover detailed community insights and sentiment summaries for Hacker News stories, reducing manual analysis time by over 80%.
Manually sifting through Hacker News comments for community sentiment is time-consuming and inefficient. This workflow automates comment extraction, AI clustering, and sentiment analysis to deliver rapid, actionable insights.
Documentation
Automate Hacker News Comment Analysis with AI Clustering
This powerful n8n workflow automates the entire process of gathering, analyzing, and extracting insights from Hacker News comments, helping you understand community sentiment and popular discussion topics with unprecedented efficiency. It's designed to transform raw comment data into actionable intelligence, saving countless hours of manual review.
Key Features
- Automated Comment Ingestion: Effortlessly pulls all comments and replies from any Hacker News story using the official API.
- Intelligent Data Vectorization: Stores comment data as searchable vector embeddings in Qdrant, enabling advanced similarity analysis.
- AI-Powered Topic Clustering: Automatically groups similar comments using K-means clustering to identify recurring themes and popular opinions.
- LLM-Driven Insight Generation: Leverages OpenAI's large language models to summarize clusters, extract key insights, and determine overall sentiment.
- Streamlined Reporting: Exports all generated insights, sentiments, and supporting comment data directly to Google Sheets for easy tracking and sharing.
How It Works
This workflow operates in two main phases: Data Ingestion and Insight Generation, orchestrated through an n8n subworkflow execution.
Phase 1: Data Ingestion Upon initiation, the workflow first clears any existing data for a specified Hacker News story_id from your Qdrant vector store. It then retrieves all comments and their replies for that story_id from the Hacker News API, flattening the comment tree for comprehensive analysis. These comments are then processed: converted into vector embeddings using OpenAI, structured with relevant metadata (comment ID, author, story title), and stored in your Qdrant instance.
Phase 2: Insight Generation Once comments are vectorized, a subworkflow is triggered. It retrieves all comment vectors for the designated story_id from Qdrant. A Python code node applies a K-means clustering algorithm to group similar comments based on their vector embeddings. Clusters containing three or more comments are then selected. For each significant cluster, the original comment content is fetched, and an OpenAI Large Language Model (LLM) summarizes the grouped comments, extracts a key insight, and determines the overall sentiment. Finally, these insights, along with supporting data, are formatted and appended to a specified Google Sheet.