NOW · FOUNDATION
Words we can trust
Establish consistent entries, validation, versioned releases, and the first authoritative language sources.
Language knowledge,
built to be taught.
Lexicon is a versioned language-data system that combines, validates, and enriches linguistic information for learning products.
Explore the architecture ↓01 · The problem
More than a dictionary entry.
Most language products start with a word, a definition, and an example. That is usable—but insufficient for deeper learning. Lexicon models words as connected, traceable language knowledge.
{
"word": "resilient",
"definition": "able to recover quickly",
"example": "She remained resilient."
}
Insufficient for structured learning
Lexeme
├── lemma
├── language
├── pronunciation
├── grammatical category
├── inflected forms
├── senses
│ ├── definition
│ ├── examples
│ ├── usage labels
│ └── difficulty
├── word family
├── semantic relationships
├── frequency
├── source provenance
└── quality status
├── lemma
├── language
├── pronunciation
├── grammatical category
├── inflected forms
├── senses
│ ├── definition
│ ├── examples
│ ├── usage labels
│ └── difficulty
├── word family
├── semantic relationships
├── frequency
├── source provenance
└── quality status
02 · Who it’s for
Designed to serve multiple audiences.
Lexicon supports several users—some current, some future. Each audience interacts with the same core language knowledge in different ways.
Current
Vocab
Lexicon provides structured learning material that powers the Vocab application—meanings, forms, pronunciation, related words, and difficulty levels.
Future
Future Veriteum language products
Shared knowledge across multiple interfaces and teaching methods, built on one consistent, versioned foundation.
Future
Developers
Versioned datasets, libraries, exports, or APIs for building language-aware applications with trustworthy data.
Future
Researchers and contributors
Source auditing, corrections, linguistic review, and dataset evaluation grounded in traceable provenance.
03 · The pipeline
From raw sources to trusted knowledge.
Lexicon follows a clear, auditable path from language data acquisition to versioned, validated releases.
01
Acquire
Import language data from approved sources—dictionaries, corpora, and linguistic databases.
02
Normalize
Convert incompatible formats into a shared, consistent data model.
03
Resolve
Separate lemmas, forms, parts of speech, and meanings into discrete, addressable units.
04
Enrich
Add pronunciation, frequency, difficulty, relationships, and contextual examples.
05
Validate
Apply automated checks to flag inconsistencies, gaps, and quality concerns.
06
Release
Publish versioned datasets for products and downstream consumers.
04 · Trust and provenance
Every field should know where it came from.
Lexicon treats source attribution as a first-class concern. Every published field should be associated with its source, license, processing history, confidence, and validation state where possible.
Sourced language knowledge should remain traceable and controlled. When a definition, pronunciation, or usage label appears in a product, the origin should be available for inspection. This is one of Lexicon’s most important differentiators.
05 · Roadmap
Lexicon direction.
The language foundation evolves across four phases—from basic structure to broad access.
NEXT · RICHNESS
Words with dimension
Add usefulness, CEFR level, pronunciation, morphology, and relationships that make each word teachable.
LATER · ACCESS
Knowledge that connects
Grow quality scoring, themed collections, meaning clusters, and learner-aware explanations.
Current maturity.
Lexicon is currently an internal language-data pipeline being developed alongside Vocab. Public datasets, developer interfaces, and contribution workflows are future possibilities, not currently available services.
See Lexicon in action within Vocab →