The purpose of this repository is to govern the creation and maintenance of metadata of processing/analysis (a/p) resources. Aspects related to the governance of codelists apply also to datasets metadata.
To create the metadata file for a new a/p resource as well as to update an existing one use the dedicated web application.
The created metadata file will automatically be published in the FAIRiCUBE Catalogue and will be queryable through Knowledge Base Query Tool.
To ensure semantic harmonisation, some metadata elements are valorised by selecting values from corresponding codelists. Use the Codelist change proposal template for proposing updates to current codelists.
NB: Each column corresponds to a codelist. The column header (in italics) is the codelist name.
| Main category | Objective | Platform | Framework | Architecture | Approach | Algorithm | Processor | OS | Use case | Conditions for access and use |
|---|---|---|---|---|---|---|---|---|---|---|
| Machine Learning | object-detection | EOX | pytorch | random forest | supervised | Random-Forest-Classifier | cpu | aix | UC1 | afl-3.0 |
| Deep Learning | classification | Rasdaman | tensorflow | DNN – Deep-Neural-Network | unsupervised | CNN - Convolutional-Neural-Network | gpu | linux | UC2 | agpl-3.0 |
| Pre-processing | segmentation | Google Colab | scikit-learn | decision-tree | semi-supervised | K-means | tpu | windows | UC3 | artistic-2.0 |
| Ingestion | regression | Kaggle | keras | ensemble | reinforcement-learning | Min-max-normalization | cygwin | UC4 | bigscience-bloom-rail-1.0 | |
| Analytics | outliers-removing | Microsoft Azure | pandas | gradient-based | DBSCAN - Density-Based-Spatial-Clustering-of-Applications-with-Noise | darwin | UC5 | bigscience-openrail-m | ||
| gap-filling | Amazon AWS | numpy | density-based | Decision-Tree-Classifier | macOS | common | bsd | |||
| feature-selection | Local Jupyter Notebook | openCV | datacubes | Random-Forest-Regression | bsd-2-clause | |||||
| dimensionality-reduction | XGBoost | RNN – Recurrent-Neural-Network | SGD - classifier – Stochastic-Gradient-Descent | bsd-3-clause | ||||||
| feature-scaling | theano | CNN – Convolutional-Neural-Network | KNN – classifier – K-nearest-neighbors | bsd-3-clause-clear | ||||||
| dataset-balancing | imblearn | Feed-Forward-Neural-Network | SegNet | bsl-1.0 | ||||||
| data-transformation | pillow | DBN – Deep-Belief-Network | LeNet | cc | ||||||
| analytics | Rasdaman | DSN – Deep-Stacking-Network | Decision-Tree-Regression | cc0-1.0 | ||||||
| clustering | MXNet | SVM – Support-Vector-Machine | Voting-Classifier | cc-by-2.0 | ||||||
| anomaly-detection | Apache-Spark | probabilistic model | AdaBoost-Classifier | cc-by-2.5 | ||||||
| gdal | Perceptron | AdaBoost-Regression | cc-by-3.0 | |||||||
| no-framework | Multilayer-Perceptron | SMOTE – Synthetic-Minority-Oversampling-TEchnique | cc-by-4.0 | |||||||
| custom-method | cc-by-nc-2.0 | |||||||||
| WCPS | cc-by-nc-3.0 | |||||||||
| Naïve-Bayes | cc-by-nc-4.0 | |||||||||
| Logistic-regression | cc-by-nc-nd-3.0 | |||||||||
| Gaussian-Mixture | cc-by-nc-nd-4.0 | |||||||||
| cc-by-nc-sa-2.0 | ||||||||||
| cc-by-nc-sa-3.0 | ||||||||||
| cc-by-nc-sa-4.0 | ||||||||||
| cc-by-nd-4.0 | ||||||||||
| cc-by-sa-3.0 | ||||||||||
| cc-by-sa-4.0 | ||||||||||
| creativeml-openrail-m | ||||||||||
| c-uda | ||||||||||
| ecl-2.0 | ||||||||||
| epl-1.0 | ||||||||||
| epl-2.0 | ||||||||||
| eupl-1.1 | ||||||||||
| gfdl | ||||||||||
| gpl | ||||||||||
| gpl-2.0 | ||||||||||
| gpl-3.0 | ||||||||||
| isc | ||||||||||
| lgpl | ||||||||||
| lgpl-2.1 | ||||||||||
| lgpl-3.0 | ||||||||||
| lgpl-lr | ||||||||||
| mpl-2.0 | ||||||||||
| ms-pl | ||||||||||
| ncsa | ||||||||||
| odbl | ||||||||||
| odc-by | ||||||||||
| ofl-1.1 | ||||||||||
| openrail | ||||||||||
| openrail++ | ||||||||||
| osl-3.0 | ||||||||||
| pddl | ||||||||||
| postgresql | ||||||||||
| wtfpl |
NB: Each column corresponds to a codelist. The column header (in italics) is the codelist name. For each codelist there are three fields on three different lines: the value, the definition and the link to a source providing more information.
| ObservableProperties | |
|---|---|
| Value | Water and Wetness |
| Definition | The HRL Water and Wetness 2018 provides primary products in full spatial resolution of 10m x 10m (as compared to 20m x 20m resolution in 2015). The main product is a classified layer, differentiating the classes of permanent water, temporary water, permanent wet, temporary wet, and dry areas, derived from water and wetness occurrences in the period 2012-2018. |
| Source Link | Water and Wetness source link |