AI4PublicPolicy Components pt.2

In our previous blog post about the main components of the AI4PublicPolicy platform, the Policy and Datasets Management and the Semantic Interoperability components have been described – find the 1st blog post on the platform components here.

In this second part of the AI4PublicPolicy platform components’ analysis, two more components from the Reusable and Interoperable Policies module are being analyzed.

Cross Country Interoperability

Component Description

Datasets and policies originate in different countries and are defined in the language of the origin country. This diminishes the chances of data and policies being reused, compared or applied in different countries. The Cross Country Interoperability component translates data from one language to another target language, so as to increase the sharing of data and policies across Europe. In particular, this component receives data in text format in a given language and produces the same information in the destination language using available translation services such as Google translate, which can be integrated with spreadsheets for translating data in tabular form.

Input

Text to be translated, language of the text and destination language.

Output

Text in the target language

Sequence Diagrams

This component takes data from a text file, the original language and the target language and invokes the translation service.

Τhe following figure shows the integration of Cross Country Interoperability with other components.

Datasets and Policies Catalogue

Component Description

The Catalogue for Policies & Datasets is a component which lists all Policies and Datasets related to AI4PublicPolicy. Information for the execution of the project is hosted in a searchable catalogue, which can be utilized by users that are in need of information related to policies and datasets for specific situations.

This searchable catalogue of policies and datasets is in the form of a registry. Through this catalogue, access to policies and data is democratized, as they are available for analysis, usage and implementation in various domains. Through the provision of such a registry, the AI tools to analyze data are only ever a click away, which leads to a reduction in the time needed for public administrators to create policies on the registered data. The catalogue also enables easy access and increased searchability to the stakeholders, by providing an adaptable search engine. The search engine offers semantic AI models, as it creates semantic links among the different keywords, for the purpose of identifying the semantic fingerprint of the searched policy or dataset and providing in real-time the requested results.

Apart from the increased searchability, the catalogue provides more knowledge base functionalities for comparison and recommendation of datasets and/or policies. By modelling datasets and policies, along with associated metadata, the catalogue can offer some decision support system functionalities, such as recommendations of related datasets or policies. This can be accomplished using a “similarity rank”, and when a user accesses a specific entry in the catalogue, it also receives information about related datasets and/or policies.

Input

This component needs as input information about the dataset or policy that should be listed, and then it retrieves and shows the information about the specific dataset and the related datasets.

Output

This component provides a searchable catalogue with capabilities to show the information about policies and datasets, but also decision support tools capabilities by providing information about related datasets or policies.

The following figure shows the interaction of the Catalogue with other components when the catalogue is browsed.