AMANDE

This is the website of the AMANDE project (1/Dec/2013 - 30/Nov/2017).

Main

Internships

No internships are available at the moment.

Internships carried out in 2018:

  • Title: Analyzing an online debate
  • Proposed by: Gabriella Pigozzi (LAMSADE).
  • Done by: Charles Nourry
  • One-line description: Public participation and online debates are increasingly encouraged. The problem is that no clear procedure to analyse, understand and summarize such debates exists.
  • More details: See the complete description here

Internships carried out in 2015:

  • Title: Influence and preferences in multi-agent argumentation systems
  • Proposed by: Stefano Moretti (LAMSADE), Gabriella Pigozzi (LAMSADE).
  • Done by: Arthur Buquet
  • One-line description: Study the different influence that arguments may have on others and how preferences hold by agents can be taken into account in the aggregation of argumentation systems.
  • More details: See the complete description here

Internships carried out in 2014:

  • Title: The power of arguments in debates
  • Proposed by: Stefano Moretti (LAMSADE), Gabriella Pigozzi (LAMSADE).
  • Done by: Benoit Pesenti
  • Description: One of the goals of this internship is to compare the relative strength of arguments in the same argumentation framework. Intertwined to this issue, is the consideration that, arguably, agents may have preferences on the arguments at stake and, in particular, may deem some arguments more important than others. Some questions that can be addressed in this internship:

• How can we define a measure of the importance of arguments in an argumentation framework? • What kind of properties such a measure should satisfy in order to keep into account both the defeat relation among arguments and the preferences of agents? The internship will mainly focus on a starting research project concerning the application of coalitional games and power indexes to the argumentation framework, and on some related issues about the computability and the implementation of such indexes.

The goals of the internships are to come up with a program allowing to: 1. generate a global argumentation framework 2. ask the user preferences about the arguments at stake 3. compute the strength of arguments based on different measures proposed recently in the literature or original methods proposed by the candidates. Note: connections are possible with the other project so as to take into account the constraints induced by the protocol in the computation of these values.