Sisyphus Academy is a multi-platform knowledge input system designed to provide ubiquitous access to personal knowledge capture through web, mobile, and browser extensions.
The current 0.1 version focuses on a Word Expansion System, where raw user input is transformed into structured, enriched knowledge entries.
Instead of functioning as a traditional vocabulary application, the system is designed as an automated data enrichment pipeline for words.
When a user submits a word, the system automatically:
- Stores the word
- Generates definitions
- Resolves or generates related images
- Builds structured metadata
- Persists enriched data into the database
This transforms simple user input into a structured knowledge entity.
User Input
↓
Word Intake API
↓
Normalization Layer
↓
Expansion Engine
├── Definition Generator
├── Image Resolver
├── Metadata Builder
↓
Persistence Layer (Database)
↓
Retrieval API
↓
Clients (Web / App / Extension)
Handles raw input from clients and ensures consistent ingestion into the system.
- Accept word input from multiple platforms
- Perform basic normalization (trim, lowercase, validation)
- Prevent duplicate entries per user scope
Transforms raw input into a standardized internal representation.
- Normalize text format
- Prepare word key structure
- Ensure consistency for downstream processing
This is the central component of the system.
Generates semantic definitions for input words.
- Can be powered by dictionary APIs or LLM-based services
- Designed to support multiple definitions per word in future versions
Output:
word → list of definitions
Associates or generates visual representations for words.
- External image search APIs or AI-based generation (future)
- Provides contextual visual enrichment
Output:
word → image_url
Constructs a structured knowledge entity from enriched data.
Includes:
- word
- definitions
- image reference
- timestamps
- tags (future extension)
Each word is stored as a structured entity:
Word Entity:
- id
- user_id
- word
- definitions
- image_url
- tags
- created_at
- User-isolated data model
- Schema designed for future knowledge graph expansion
- Optimized for retrieval and enrichment workflows
Provides access to enriched word data.
- Word list retrieval
- Single word lookup
- Filtering support (future)
- Learning-oriented query extensions (planned)
The system is designed to support multiple input sources:
- Web application
- Mobile application (planned)
- Browser extension (primary input interface)
All platforms share a unified backend processing pipeline.
- Word-centric data model design
- Expansion pipeline architecture definition
- Backend system structure setup
- Expansion Engine implementation
- External API integration (definition/image)
- Persistence optimization
- Word ingestion
- Automated definition generation
- Image resolution
- Structured storage
- Example sentence generation
- Multi-definition support
- Spaced repetition system
- Difficulty estimation
- Personalized word recommendations
- Learning pattern analysis
- Adaptive review scheduling
- Expansion from words to concepts
- Connection between words and notes
- Knowledge graph structure
- Cross-domain semantic linking
The system is built on the principle that:
Input should be immediately transformed into structured and enriched knowledge.
The goal is not simple storage, but automatic transformation of raw input into usable knowledge structures.
- Java 21
- Spring Boot 3.4+
- Spring Security (OAuth2 + JWT)
- PostgreSQL
- Redis
- Gradle
- H2 (testing)
- Multi-platform input support
- Automated data enrichment pipeline
- Extensible knowledge schema
- Separation of concerns across pipeline stages
- Designed for future AI integration
Sisyphus Academy v0.1 is defined as:
A system that transforms raw word input into structured, enriched knowledge entities through an automated expansion pipeline.
./gradlew clean build
./gradlew test
./gradlew bootRunApache License 2.0