Generate content, correct spelling, suggest synonyms to enhance writing.
EpubGraph
0Popularity
About EpubGraph
EpubGraph is a local application for managing and exploring ebook libraries. It processes EPUB files to extract metadata, generate AI-powered summaries, and create semantic embeddings that map relationships between books. The tool uses an 8-billion-parameter model to produce 2048-dimensional vectors representing each book’s themes, tone, and style. These embeddings feed into a hybrid recommendation algorithm that combines content similarity, author links, series connections, and user ratings to suggest new reading material. A force-directed graph visualizes how books relate through shared topics, authors, or series, while a 3D bookshelf renders the reading queue with realistic lighting and textures. All operations run locally on the user’s device, with no cloud dependency or telemetry. The application integrates with Calibre libraries and supports keyboard shortcuts for efficient navigation and management.
Key features
- AI-generated book summaries
- Semantic embeddings of 2048 dimensions
- Hybrid recommendation algorithm
- Interactive book graph visualization
- 3D bookshelf with WebGL rendering
- Calibre metadata import
- Local file pointer storage
- Keyboard shortcuts for navigation
Use cases
- Discovering new books based on thematic connections
- Managing large ebook collections with semantic search
- Visualizing relationships between books in a personal library
Pros
- Local processing with no cloud dependency
- Semantic embeddings for contextual book relationships
- 3D interactive bookshelf visualization
- Calibre library compatibility
- Privacy-focused with no account requirements
Cons
- Requires Ollama for AI inference
- No mobile or web version
- Limited to EPUB format
- Initial setup involves model downloads
Frequently asked questions about EpubGraph
What is EpubGraph and what does it do?
EpubGraph is a local desktop application designed to help users manage and explore large ebook libraries. It processes EPUB files to extract metadata, generate AI-powered summaries, and create semantic embeddings that map relationships between books.
Who is EpubGraph suitable for?
EpubGraph is ideal for avid readers with extensive ebook collections who want to discover new books based on content similarity, author links, series connections, and personal ratings. It is also useful for users who prioritize privacy and prefer local processing over cloud-based solutions.
How does EpubGraph generate book recommendations?
EpubGraph uses a hybrid recommendation algorithm that combines content similarity, author relationships, series connections, and user ratings. It employs an 8-billion-parameter model to create 2048-dimensional embeddings for each book, which are then used to identify meaningful connections and suggest new reading material.
Does EpubGraph require an internet connection to work?
No, EpubGraph operates entirely locally on the user’s device. It does not rely on cloud services, telemetry, or external accounts, ensuring that all data remains private and stored on the user’s machine.
How do I get started with EpubGraph?
To get started, download and install EpubGraph, then point it to your ebook folders or Calibre libraries. The application will automatically extract metadata and generate AI-powered summaries and embeddings. Ensure Ollama is installed and running to handle the local AI inference.
Can EpubGraph integrate with Calibre libraries?
Yes, EpubGraph is compatible with Calibre libraries. It automatically detects Calibre metadata.db files and imports metadata, ratings, tags, and covers non-destructively, allowing seamless integration with existing Calibre setups.
