We are delighted to share Climate Policy Radar’s Public Product Roadmap – sketching the steps we’re taking today to reach our (very) ambitious future goals.
Transparency is one of our core values, and it threads through everything we do and advocate for. We believe that being open about our own processes and priorities will lead us to make decisions that are rooted in the needs of our users and stakeholders. It will enable us to move, iterate, fail, and retry faster, in the face of the growing climate emergency.
Our roadmap is both practical and aspirational. You can use it to track the progress we’ve already made on developing our research tool, what we’re working on now, and what we plan to work on in the future. We’ve set our sights high.
Alongside charting the direction we’re taking to achieve our goals, we’ve written a brief description under most items on the roadmap so you can learn more about what the work involves.
Organising, analysing and democratising data
Our goal is to provide data and tools to support better decision-making. Robust, accessible data equips and empowers policymakers to learn from successes and mistakes and design better policy responses. Data helps track and evaluate collective climate action, as well as contributing to policy risk analyses for mobilising public and private finance. Evidence also helps citizens and civil society hold decision-makers to account through democratic engagement and action.
Our work is centred around three strategic pillars that aim to address these systemic challenges: organise, analyse and democratise data. The items on our roadmap are segmented into our broad areas of work against these goals – traversing from left to right on the map will guide you through what we’ve built all the way to what we’re dreaming about. Below we’ve mapped out our broad plans across these pillars.
1: Bringing in more datasets and documents (on roadmap as “New Data”)
We build data pipelines that ingest and translate documents, bringing them onto a single platform to make them easily discoverable and comparable.
On our immediate horizon (this year) is widening our data offering by bringing in a range of other datasets. This includes documents from the UNFCCC (for example, NDCs and documents submitted under the first Global Stocktake); climate litigation documents from our partners at the Sabin Center; disaster risk reduction policies (with UNDRR’s Preventionweb); policies about nature-based solutions (with Open Earth Foundation and Arboretica), and more. Beyond that, we’re dreaming of collating state- and city-level documents, and expanding beyond climate to other closely intertwined domains like biodiversity and health.
2: Using AI to derive insights from documents (on roadmap as “Data Science and AI”)
Law and policy documents are often long, and dense with data and insight. Because of this, they can be inaccessible and opaque. So as well as making documents accessible, we work to make them useful for policy tracking, evaluation and analysis by extracting structure, meaning and insights from them. To do this we’re applying machine learning and natural language processing to identify key concepts in the text – for example, the technologies a policy is seeking to accelerate or phase out, the emission reduction targets a country has committed to, or the policy instruments it employs.
These concepts are the foundation on which we’re building a knowledge graph, which will string together the concepts identified, and represent a network of connections and relationships between entities, like extreme weather events and adaptation measures aimed to address those. This opens up the ability to explore policy data in detail, and eventually, extend and connect it to other datasets – like budgets allocated, emissions, deforestations metrics and more – to start answering the big question of – what impact do these policies have?
We are also building research tools to interrogate the data, including exploring how we can leverage large language models such as BERT and OpenAI’s ChatGPT and GPT-4 as virtual research assistants. This includes summarising and comparing text, or generating content maps of documents, as well as data visualisations, time series creation and more.
3: Enhancing our user interface (“User Interface”)
Our research tool includes country profiles, semantic search that saves time by recognising and looking for similar phrases to your search term, and a highlighting tool that shows you where your search term or related words appear in the text of documents – no ‘CTRL+F’ needed.
We have many more things in the pipeline to help people get more out of our data. We will make key concepts identified by our data scientists easy to find in a document (eg. ‘show me all the renewable technologies in this document’). Documents from all languages will be automatically translated to English, and longer-term the entire user interface will be made available in other languages too (probably starting with French and Spanish).
We will find new ways to allow users to visualise and explore our data, and make structured data available via APIs for others to use. We are thinking of personalised features like saved searches and email alerts.
The creation of our law and policy knowledge graph will enable new interfaces for exploring policy insights, and will combine with our experiments with large language models (like ChatGPT) for guiding our users to the information they need.
This roadmap is…
What we are working on now, and what comes next: We’ve divided this into a few rough time blocks (Live, Building, Designing, Planning, Dreaming) to signal when we anticipate working on each task.
Focussed on data and product changes that are user-facing: We don’t include all the under-the-hood functionality for internal users like developers or internal admin systems (for example, policy analysts who add and edit data).
A living document: You can visit the link at any time to view the latest version. We’ll also share how our roadmap is progressing and evolving through blog posts and newsletters – sign up to receive our newsletter. If you’re keen to know more about what we have already done, follow Climate Policy Radar’s Changelog that details our latest product developments.
This roadmap is not…
Exhaustive: We’re coming up with new ideas all the time, so check back in regularly to see our latest ideas and plans.
Set in stone in terms of content or timeline: There is increasing uncertainty as we traverse the roadmap. This is because as we continue to learn and grow, we’ll generate new ideas, respond to feedback, push the boundaries for what’s possible – and also, no doubt, bump against unforeseen challenges and hurdles.
Tell us your thoughts
We’d love to hear feedback both on what our users and stakeholders need – and on how we might go about meeting those needs, from a technological perspective. If you want to share your ideas, wish lists (and dreams!), or any other feedback on this roadmap please get in touch. We’d especially be grateful if you’d be happy to participate in a short user interview with our product manager.
This blog post was originally published on Climate Policy Radar