The ManyLaws Project (EU-wide Legal Text Mining using Big Data Processing Infrastructures) is a platform set up to deliver a set of services for citizens, businesses and administrations in the European Union, Austria and Greece.
The Action's overall objective is to enable access to legal information across the European Union and improve the efficacy of decision making in legislative procedures operated by public bodies. Specifically, the Action will set-up a platform (ManyLaws) which will deliver a set of services for citizens, businesses and administrations built upon text mining, advanced processing and semantic analysis of laws of the European Union, Austria, and Greece. The addition of laws from other Member States will be explored and, if possible, implemented. Moreover, these services will be tested within at least two law making procedures of the Greek and Austrian Parliament.
The services that will go-live by the end of the Action are:
- Exploring national legal corpora in English, Greek, and German
- Exploring the alignment of the Austrian and Greek legislation with the EU legislation e.g. assessment of the degree of transposition of EU Directives into national law
- Comparing national Austrian and Greek laws which target the same life events
- Parallel search in the Austrian and Greek legal frameworks through parallel translation of search terms, and using simple keywords
- Analysis of references to the European legislation by national laws and within the same Member State
- Comparative analysis of related laws within the same Member State
- Timeline analysis for all legal acts, visualising the progress and current status of a specific national or European piece of legislation over time starting from preparatory acts until its adoption and including amendments and extensions. This service will help users to follow the procedures that have led to the existing legal act
- Visualizations of correlations, dependencies and conflicts between different laws, including geo-related visualisations, text-related visualisations
- Decision support service for legal procedures. The service will help public stakeholders to identify laws that are affected or contradicted by a prospective policy
- Interrelation of laws and news or social media posts, including sentiment analysis
To achieve this, the Action will combine data on laws produced by the EU, and at least by Austria and Greece from 1990 to 2020; annotate it; link it and transform into structured open datasets. This process will be enabled by the use of supercomputing resources. Data covered will include: law titles, subjects, publication reference, dates of publication, keywords, article numbers (structure), and references to other laws. The majority of data will be in textual format, while numeric information will also be extracted. Based on the metadata that will be identified a new data model will be described that is compatible with the European Data Portal.
The service portal that will be delivered by the Action, will be designed and built based on the requirements to be harvested by the Open Data Greece portal of the Ministry of Administrative Reconstruction and by the Open Government Data portal Austria. Specifically, the metadata will be encoded using the W3C Data Catalog Vocabulary (DCAT) that will allow the populated dataset to be uploaded on the above mentioned portals, which are harvested by the European data portal (EDP). In the event that publication on the two aforementioned portals is not feasible, for reasons beyond the beneficiaries control, ManyLaws will at least ensure that datasets will be permanently made publicly available as open data also after the duration of the Action.