The text labels and filter tools in the Global Stocktake Search are based on data collected using Explorer, a text analysis tool developed by Climate Policy Radar. Explorer enables search for references to a number of concepts simultaneously across a large number of documents, using keywords. This section documents how Explorer works, and the methodology we follow to develop keyword lists in order to identify references to concepts within documents related to the Global Stocktake.
To search a set of documents for references to a concept using Explorer, we create bespoke CSV files for each concept, containing keywords organised across a three-tier hierarchy. For example, for fossil fuels:
| Tier 1 concept | Tier 2 concept | Tier 3 concept |
|---|---|---|
| Fossil fuels | Oil | Crude oil |
| Fossil fuels | Oil | Petroleum oil |
| Fossil fuels | Oil | Gasoline |
| Fossil fuels | Gas | Natural gas |
| Fossil fuels | Gas | Petroleum gas |
In our input files, we indicate where a certain keyword may have multiple synonyms or different linguistic expressions, such a present/past tense and singular/plural. All of these variations are then searched for automatically. This enables us to perform a precise and comprehensive search for each concept without exhaustively listing all of its possible linguistic expressions.
Wherever possible, we aim to rely on existing taxonomies developed by expert third-party sources as the basis for our keyword lists. For example, our keyword list for technologies is based on the UN Climate Technology Centre and Network’s Taxonomy of Climate Technologies, while our keyword list for climate-related hazards is based on the UN Office for Disaster Risk Reduction’s Hazard Definition and Classification Review.
Where such resources are available, we take the following steps to develop a keyword list:
- We evaluate existing resources to assess whether they are developed by credible sources (for example, a UN agency) and whether they are widely used and referenced.
- We use the taxonomy to create keyword lists following the process described in Section 2 above.
- We use desk research to expand these keyword lists and ensure that they include as many relevant terms as possible.
Where we are not able to identify an existing credible source providing a taxonomy for a given concept, we create keyword lists from scratch based on our own internal expertise and on desk research.
This process involves reading material from sources such as Wikidata, as well as academic journals and reports and publications by international agencies, to identify the relevant key terms for a given concept.
Based on this desk research, we create keyword lists following the process described in Section 2 above.
When performing a pre-set search using the filters on the left side of the page, you will see that phrases in the text of a Global Stocktake document that represent or express a concept are tagged and labelled.
Use these labels to identify instances of your selected concept, and read the surrounding text to understand the context in which that concept has been referred to in the document.