1 Roadmap

Stages in the roadmap are often iterative and will be repeated, revisited, refactored as appropriate.

Key

AR6 - IPCC Sixth Assement Report

CPS - Computational Publishing Service

#semanticClimate - project partner - open research group building software tools for climate knowledge liberation

Convert AR6 report corpora web source to normalised HTML (web scrape) - v1.0 Ready
Locate supporting resources located: Authors, glossary, acronmys, figures, data, bibliographic, licence, published canonical sources, references - v1.0 Ready
Knowledge graph data model - Alpha
LLM RAG workflow tests - Alpha
Publishing pipeline from Wikibase - Beta 1.0 Ready - See python module: https://pypi.org/project/cps-impress/
AR6 Report quantification - v1.0 Ready (repeat using TIB Grobi https://projects.tib.eu/grobi )

Wikibase infrastructure - WIP
AR6 Report publication sytactic, semantic, and typesetting data stucture analysis - WIP
AR6 Report publication CSS Paged Media style - WIP
Defining workflow and tech stack - WIP
Wikibase maintenance and sysadmin - WIP

AR6 WG1 section trial import into Wikibase for the purpose of Wikibase troubleshooting
Wikibase text import module - A python module needs writing importing text using Wikibase and a storage method - Renate or/and Mediawiki. This would extend https://pypi.org/project/cps-wb/
Refector AR6 web scrape - A python module needs extending. This module was made by CPS for an art collection https://pypi.org/project/cps-deckenmalerei/
Import AR6 into Wikibase
Harvest data: Python module needed. This would extend https://pypi.org/project/cps-wb/
- Authors
- Glossary
- Acronyms list
- References
- Bibliographic
Import above data to Wikibase. Python module needed. This would extend https://pypi.org/project/cps-wb/
Annotate report using above data
Wikibase to Mediawiki report navigation mapping for report browsing

Define workflows and review processes for adding data to KG.
LLM RAG proof of concept demo - using HTML corpora and Streamlit. This is meant as a throw away demonstration as the workflow has been tried out with #semanticClimate
Community annotation (integrate #semanticClimate tooling): #semanticClimate, Stockholm Climate Institute (SEI), Potsdam Climate Institute (PIK), UNESCO, UNFCCC, etc.
Wikibase to Wikidata data mapping
Wikibase data analysis and visualisations - using built in Wikidata tools
Publications available as:

REST API
Command line
Jupyter Noteooks
Python CMS (e.g., Wagtail)
Graph RAG LLM
- All of the above publishing channels use the following framework. Computational Publishing Service (CPS) using the publishing engine (CPS_Impress). CPS_Impress publishes from Wikibase to HTML using Jinja templating in a ‘Model View Controller’ architecture. Paged Media CSS styles are used to create PDF like layouts. Publications are saved back to the knowledge graph and online as sharable resources.