All notable changes to this project will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Project structure to match the structure of projects in OpenStack.
- The search has been re-enabled. It had been inadvertently commented out.
- Issue with search results state not being initialised.
- RSECon 2020 poster and slide.
- Issue with the latest version of Poetry (1.2.0) failing to install fasttext 0.9.2.
- Project documentation.
- Filtering was being applied even when no filters were selected, this was preventing environment statements from displaying because a lot of them don't have filter values.
- Colour the parallel diagram by topic group.
- Colours match between the different topics visualisations.
- Increase the default topics threshold to 0.75.
- Filtering by field of research could produce confusing results because it included documents without a FoR.
- Explanation about text summarisation.
- Display selected document under the about the data section.
- Update Streamlit and txtai to latest versions.
- Sidebar labels.
- Text search filter interface.
- KDL footer.
- Extract unit of assessment panel.
- Panel filter to the dashboard.
- Entity type glossary.
- Entities filter.
- UK nations distribution.
- Command to extract strengths, areas for improvement and future plans.
- Rename types of impact to outputs.
- Sort units of assessment by number.
- Group types of output.
explain
module to explain the outputs of the classification models.- Entity type filters to the dashboard.
- Topic based filters to the dashboard.
- About the data section to the dashboard.
- Colour the parallel diagram by topic group.
- Colours match between the different topics visualisations.
- Increase the default topics threshold to 0.75.
- Filtering by field of research could produce confusing results because it included documents without a FoR.
- Explanation about text summarisation.
- Display selected document under the about the data section.
- Update Streamlit and txtai to latest versions.
- Sidebar labels.
- Text search filter interface.
- KDL footer.
- Extract unit of assessment panel.
- Panel filter to the dashboard.
- Entity type glossary.
- Entities filter.
- UK nations distribution.
- Command to extract strengths, areas for improvement and future plans.
- Rename types of impact to outputs.
- Sort units of assessment by number.
- Group types of output.
explain
module to explain the outputs of the classification models.- Entity type filters to the dashboard.
- Topic based filters to the dashboard.
- About the data section to the dashboard.
- Product to the list of entity types to extract.
- MPL-2.0 license file.
- Tests for
etl
module. - pre-commit configuration.
- Processing for environment studies.
- Help information to the dashboard.
- Geo data to count/display only unique mentions per case study.
- Group places by country to display data in the map.
- Move all the chart creation to the visualize module.
- Enable click selection in the data grid.
etl
module to extract data from environment statements.- Reduce the number of columns in the data grid.
- Refactor sidebar options.
- Dynamically calculate the plot heights.
- Move text data to the
y
axis for easier readability.
- Person from the list of entity types to extract.
- Labels being cut-off in parallel plots.
- Bubble map to the dashboard.
- Module to create visualisations.
- Streamlit cache to use the new recommended method of caching.
- Chart to show correlation between topics and units of assessment.
- Docker compose files.
- Extract research section to be used to extract fields of research.
- Streamlit configuration file.
- Alluvial diagram to show relationship between topics and units of assessment.
- Country information to the place entities.
- Country related charts.
- Persistent cache for geocoding.
- Mapbox token.
- Topic classification to support extracting from different sections and using different labels.
- Enable multi-label topic classification.
- spaCy language model to a transformer based model for better accuracy.
- Geolocate feature to return GeoJSON alongside with the data.
- Command and module to extract, transform and load data.
- Text summarisation.
- Topic classification.
- Dashboard with data, summary and topics chart.
- Entity extraction.
- Entity geolocation.