/Personal-Browsing-Embeddings-Extension

Personal Browsing Embeddings Extension

This browser extension enables users to capture and encode web content or selections, creating a personal embeddings database. The extension features a chatbot for searching the database, finding similar content, and using language models (e.g., GPT-4) for analysis. Use it to efficiently access and interpret past research.

Table of Contents

Features

  • Scrape and encode web page content or highlighted sections
  • Store embeddings in a personal database
  • Tag stored embeddings by project or topic
  • Query the embeddings database using a chatbot interface
  • Perform similarity searches within the database
  • Leverage language models (GPT) for content interpretation and discussion

Optional Features

  • Always-On Encoding: Constantly record all content during normal browsing. (Probably a bad idea.)
  • History Encoder: Encode content from links referenced in browser history.
  • URL Encoder: Scrape and encode content from user-specified URL or list of URLs.
  • Vector Database Options: Choose between local or hosted vector database service (pinecone.io).
  • Encoder Options: Use open source (txtai) or API-based encoder (openai).
  • Retrieval Only vs. LLM Interpretation: Choose retrieval only or language model interpretation.
  • Selective Deletion: Delete specific embeddings and associated metadata.

User Experience

Use the extension to efficiently capture, encode, and retrieve web content:

  • Installation:

    • Install the extension from your browser's extension marketplace (e.g., Chrome Web Store, Firefox Add-ons).
    • An extension icon appears in your browser's toolbar for easy access.

  • Content Encoding:

    • Highlight text or navigate to a page, and click the extension icon to encode and store content.
    • The encoded content and metadata (e.g., link reference) are stored in your database.

  • Chatbot Search:

    • Use the chatbot to search your database for content related to your query.
    • The chatbot displays matching results and links to original sources.

  • Language Model (Optional):

    • Enable language model interpretation for comprehensive summaries and insights.
    • The chatbot passes your query to a language model (e.g., GPT-4) for interpretation.

  • Database Management:

    • Control, organize, and delete content in your database through the extension's interface.

Example Use Case: Art History Student

As an art history student researching Renaissance painters, use the extension to capture and retrieve information:

Recording Research Content

  • Use the extension to capture content from online sources about Renaissance painters.
  • Store the content and metadata (e.g., link references) in your database.

Remembering Details and Gaining Insights

  • Query the chatbot for details about Leonardo da Vinci's painting techniques.
  • Get matching results, original source links, and language model interpretations.

Advancing Study Goals

  • Make connections between painters, techniques, and movements with easy access to information.
  • Use insights for class discussions, essays, and presentations on Renaissance art.

The extension supports your academic research and helps you gain valuable insights.

Development Steps

  • Browser Extension:

    • Create an extension compatible with browsers (e.g., Chrome, Firefox).
    • Enable scraping of web page content or user-selected text.
    • Add a user interface with an activation button.

  • Embeddings Encoder:

    • Select an embedding model (e.g., Word2Vec, BERT) and decide on pre-trained or custom embeddings.
    • Implement encoding to convert content into embeddings.

  • Database:

    • Design a database for storing embeddings, metadata (e.g., link references), etc.
    • Decide on database location (e.g., local storage, cloud) and ensure security.

  • Chatbot and Search:

    • Build a chatbot interface for user queries.
    • Implement similarity search to find related content in the database.
    • Integrate an external language model (e.g., GPT-4) for query interpretation.

  • User Experience and Privacy:

    • Create a user-friendly interface for installation, database management, and chatbot interaction.
    • Address privacy concerns and obtain user consent for data usage.

  • Testing and Deployment:

    • Test the extension for functionality, performance, and usability.
    • Fix bugs and issues found during testing.
    • Deploy on browser extension marketplaces (e.g., Chrome Web Store, Firefox Add-ons).

  • Documentation and Support:

    • Provide documentation for installation, usage, and troubleshooting.
    • Offer support channels (e.g., email, forum) for user assistance and issue reporting.

Contribute

Explore the repository, learn about the extension, and contribute to its development. Thank you for your interest in the Personal Browsing Embeddings Extension!