Lavoisier is a language which allows the selection of entities from a domain model and transforms the selected entities instances into tabular-formatted data, which can be used as input of data mining algorithms for an analysis.
You can find data selection examples in the lavoisier-example Eclipse project of this repository, or in the lavoisier-evaluation external repository where a comparison between Lavoisier and state-of-the-art technologies for data selection and formating is shown.
Xtext 2.8.4.
- Import "es.unican.lavoisier..." projects into an eclipse workspace.
- Generate model code for all genmodels present in es.unican.lavoisier.domainModels/models/
- Right-click "es.unican.lavoisier/src/es.unican.lavoisier/Lavoisier.xtext" and run "Generate Xtext artifacts".
- Right-click "es.unican.lavoisier" project and select "Run Eclipse Application".
- In the newly opened eclipse instance, import "lavoisier-example" project.
- The file "extractions/dummy.lv" inside that project is a simple example of dataset specifications over a domain model. CSV files are generated at src-gen folder.
Pinset is a language that follows similar principles and objectives to those of Lavoisier. It has been implemented on top of the Epsilon platform, so its main focus is to provide a modelling tool for software engineers to extract datasets from models in a model-driven engineering context.
@article{lavoisier2020,
author = {Alfonso de la Vega and
Diego Garc{\'{\i}}a{-}Saiz and
Marta E. Zorrilla and
Pablo S{\'{a}}nchez},
title = {Lavoisier: {A} {DSL} for increasing the level of abstraction of data
selection and formatting in data mining},
journal = {J. Comput. Lang.},
volume = {60},
pages = {100987},
year = {2020},
url = {https://doi.org/10.1016/j.cola.2020.100987},
doi = {10.1016/j.cola.2020.100987}
}