skills.rest
Telepotch
TelepotchOFFICIAL
Explore all skills by this author →
GitHub

edi-lexicon

Normalize sampling data terms into a canonical LEXICON.jsonl with decisions.json.

Updated Aug 27, 2026
Search Tags:#sampling#lexicon#terminology#editorial-workflow#text-normalization#ner#term-dictionary
One-click install
npx skills add https://github.com/Telepotch/hirano-edi-on-claude-code --skill edi-lexicon
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: edi-lexicon
Source: https://github.com/Telepotch/hirano-edi-on-claude-code/tree/main/.claude/skills/edi-lexicon
Command: npx skills add https://github.com/Telepotch/hirano-edi-on-claude-code --skill edi-lexicon

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires neologdn, jaconv, spacy, ginza, ja-ginza, sudachipy, and includes scripts (resource) components.

What problem does it solve?

This skill extracts and refines project-specific terminology from sampling data to produce a consistent, auditable LEXICON.jsonl.

Core Features & Use Cases

  • Extracts named entities and term candidates from JSONL sampling data
  • Applies editorial rules to unify variants, merge surnames, and filter noise
  • Generates a finalized LEXICON.jsonl suitable for downstream processing and reviews

Quick Start

Run the lexicon build pipeline on your current project data.

Frequently Asked Questions about edi-lexicon

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a project-specific lexicon from sampling data?▼

This Skill extracts named entities and term candidates from JSONL sampling data to build a consistent lexicon. It normalizes text, unifies variants, and filters noise to produce a finalized LEXICON.jsonl output.

How does named entity recognition work for terminology extraction in an editorial workflow?▼

Named entity recognition for terminology extraction uses spaCy and Ginza to identify term candidates from JSONL sampling data. The pipeline applies editorial rules to unify variants, merge surnames, and filter noise into a canonical LEXICON.jsonl.

Can I use spaCy and Ginza for Japanese text normalization and term consolidation?▼

Yes, spaCy and Ginza alongside neologdn and jaconv handle Japanese text normalization and term consolidation. This Skill leverages these dependencies to extract, normalize, and consolidate project-specific terms from sampling data.

What is the best way to review and approve term dictionary changes?▼

The best way to review term dictionary changes is to generate a decisions.json file for auditing. This Skill outputs a decisions.json for review and a finalized LEXICON.jsonl reflecting approved changes while preserving high-confidence existing entries.

Does this lexicon build pipeline preserve existing high-confidence term entries?▼

Yes, the lexicon build pipeline preserves high-confidence existing entries during consolidation. It identifies and normalizes new term candidates from sampling data while maintaining validated entries in the finalized LEXICON.jsonl.

Why does my terminology extraction output include typos and surname variants?▼

Terminology extraction outputs include typos and surname variants because raw sampling data contains noise. This Skill applies variant handling, surname merging, type corrections, and typo filtering according to predefined editorial rules to clean the finalized LEXICON.jsonl.

Related Skills

Explore More Content & Communication Skills→
robertgussrobertguss
103

writing-dna-discovery

Creates a Voice DNA Document from collaborative interviews and writing samples.

Community
Advanced
TheMattBermanTheMattBerman
226

google-ads-apply

Execute approved Google Ads draft actions with safety checks and audit trails.

Community
Advanced
48Nauts-Operator48Nauts-Operator
3

langchain-architecture

Design LangChain LLM applications with agents, memory, and tool integration.

Community
Advanced
tonombtonomb

ubiquitous-language

Extract domain terminology and definitions into a Markdown glossary.

Community
Intermediate