• RANKED
  • FILTERED
  • HUMAN
  • SUMMARY

How to rank documents by keyword relevance with your AI

Connect this MCP to your AI client to turn raw text files into a sorted, searchable library. Once connected, your agent handles the math to find exactly what you're looking for.

Ask AI about this page

Short answer

How do I sort my files by how much they talk about a specific word?

Your AI uses the TF-IDF Vectorizer Engine to calculate a mathematical score for every document in your set. This score tells you which files are actually relevant to your keyword versus those that just mention it passingly. You'll get a ranked list of files ordered from most relevant to least relevant.

Where every result lands.

The outcomes of your search.

The engine processes your files and organizes them into these specific views.

RANKED

Relevance order

The AI produces a list of files sorted by their TF-IDF score. You get the most important documents at the top.

FILTERED

Keyword matches

The engine isolates only the files that meet your specific term threshold. This clears out the noise from your collection.

HUMAN

Manual review

You decide which high-scoring documents are actually useful for your specific context. The AI provides the math, but you provide the judgment.

SUMMARY

Topic overview

The AI identifies which terms are most unique to your document set. This helps you understand the core themes of your data.

The workflow

What your AI does when the files arrive.

The AI moves from raw text to a structured, mathematical ranking.

  1. Analyze text

    The agent reads through your provided text files or articles to understand the word distribution.

    calculate_tf_idf

  2. Calculate scores

    The engine runs the TF-IDF algorithm to determine how much each term matters to each document.

    calculate_tf_idf

  3. Rank documents

    The AI sorts the entire collection based on the calculated relevance scores.

    calculate_tf_idf

  4. Present results

    Your AI client shows you the final list, highlighting the most relevant files first.

    calculate_tf_idf

Try it

Copy these to start.

Use these prompts once you've connected the MCP.

Starting points

These are starting lines. Replace the bracketed text with your actual folder names or specific keywords.

natural Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_CzteX5B7baA5kKcgQISGvHapYHeKBksKMIECPMUI/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — natural capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "tf-idf-vectorizer-engine-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_CzteX5B7baA5kKcgQISGvHapYHeKBksKMIECPMUI/mcp"
    }
  }
}
Copy into chat04

  • Rank the documents in my 'Project_Alpha' folder by how much they discuss 'budgeting'.

  • Look through the articles in my 'Research_Notes' directory and find the most relevant files for the term 'quantum computing'.

  • Sort my 'Client_Feedback' folder by relevance to the keyword 'user experience'.

  • Which files in my 'Legal_Docs' folder are most heavily focused on 'compliance'?

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect TF-IDF Vectorizer Engine once, then ask.

Just click the link to connect the MCP to your client. Your credentials stay encrypted, and you can start ranking files immediately.

Connect TF-IDF Vectorizer Engine to your AI

FAQ

How this task behaves.

  • 01

    Can the AI rewrite my documents to make them more relevant?

    No. This MCP only reads and calculates scores for your existing files. It does not modify your content.

  • 02

    Does the AI delete files that have low relevance scores?

    No. The AI only ranks and sorts them. Your files remain exactly where they are.

  • 03

    How does it handle files that don't contain the keyword at all?

    Files without the keyword will receive a score of zero and appear at the bottom of your ranked list.

  • 04

    Can I use this to search through my email inbox?

    This specific tool works on text files and articles you provide to the agent. It does not have a tool to access your email directly.

  • 05

    Is the ranking based on meaning or just word counts?

    It uses TF-IDF, which is a mathematical way to weigh how unique a word is to a specific document compared to the whole set.

  • More questions about TF-IDF Vectorizer Engine? The Connector page answers them. See everything the TF-IDF Vectorizer Engine Connector can do

Connect natural to Claude, Cursor, ChatGPT & more