Skip to content

Latest commit

 

History

History
48 lines (25 loc) · 2.6 KB

README.md

File metadata and controls

48 lines (25 loc) · 2.6 KB

AI4Lobbying - Building AI course project

AI project to help ethical lobbying

Summary

AI4Lobbying collects all available data from EU institutions on a specific subject. It establishes links between Members of the European Parliament and their expressed votes to create a map of interests.

Background

The idea tackles the issue of identification of supporters, or not, of a specific idea. It is a key issue for lobbysts to be able to identify the two camps, to have the possibility to explain and motivate the position.

How is it used?

Any interest group, be it representing environmental interests, consumer insterests or sectorial interests, has to identify the camp of supporters and the opposite camp. Ideally, also the camp of "not yet decided" would be useful, so to possibly convince them. It is THE problem to tackle at every new legislative proposal. It would greatly improve the available knowledge. By expressing explicitly each other interests, cross party and cross nationalities, it woudl also enhance the level of transparency of the political process.

The Ai4Lobbying would be the ideal tool to map the situation and dinamically improve it at every step in the legislative process.

Data sources and AI methods

Data comes from all available European Institutions public websites, for instance:

https://www.europarl.europa.eu/portal/en

https://ec.europa.eu/info/index_it

https://www.consilium.europa.eu/en/european-council/

Corroborated by analysis done in online publicly available news magazines, for instance:

https://www.politico.eu

https://www.endseurope.com

The AI methods used would have to categorise the data found, itemise them, label them according to being "in favour" or "opposed" to the key idea. It would also look into proposed amendments to reinforce, or challenge, the map of interests built.

Challenges

AI4Lobbying does not identify the strategy to follow, it just maps the situation as it is. By trying to identify supporters of a given idea, could also be used to monitor the way in which each Member of the European Parliament position himself with respect to specific political issue. While the aim is to increase the transparency in the political discussions, any possible misuse or abuse of the information should be avoided.

What next?

The evolution of AI4Lobbying would also identify a possible strategy in support of the chosen interest group. That would require more knowledge of the system and more powerful learning algorithms.

Acknowledgments

Many thanks to my colleagues, and opponents, in everyday work. All of them would benefit from the AI4lobbying - equally.