[AT workshop] Session 2

On the use of argumentation in agreement technologies
(Henry Prakken)

Agents need argumentation (i) for their internal reasoning and (ii) for their interaction with other agents. Explaining basic things about argumentation process: argument attacks. The situation of the dialog can be modeled in a graph colored by defining in and out arguments (Dung, 1995). And there is a sound and complete game that allows to determine if an argument is feasible or not without having to calculate the entire network: an argument A is feasible when there is a winning strategy for A follow.ing the game rules.

Problem: it is asumed that all information is centralized and static (a single theory -KB-) So dialogue game systems are developed. He’s using the Walton & Kreebe dialogue types (without eristic :-) I’ve seen this a lot of times already.

An Interesting thing: blocking behavior (always asking why) It can be solve by using sanctions:
social sanctions (i wont talk you any more)
shift of burden of proof by a third party (referee): q since r // why r? // referee: you must defend not-r

I already knew most of these things (thanks to Stella)

«Prof. Kripke, let me introduce Prof. Nash», or
Logic for Automated Mechanism Design
(Mike Wooldridge)

In MAS the interaction is done by mechanisms = protocol + self-interest and agents are the participants in these mechanisms. So mech. can’t be treated as simple protocols. (ex. sniping in eBay -bidding in the last 5 min. trying to be the last bidder-). A MAS can predict the sniping behav. of users in eBay?  The environment of an agent is a mechanism too, that contains other agents that act strategically to achieve their own goals

The formalization used is ATL (alternate-time logic), introduced in 1997 top analyze games. It defines a branching-time model as a graph and CTL is the logic used to talk about branching-time structures, extending propositional logic with path quantifiers (A,E) and tense modalities (F, G, X, U).

CTL sais when something is inevitable or possible, but it hasn’t notion of strategy action nor agency (it’s a problem to model mechanisms…. and service-based applications too). ATL is intended to overcome these limitations. The basic expression is

\(\langle \langle C \rangle \rangle \phi\)

meaning «coalition C can cooperate to ensure that \(\phi\). The idea is that, using coalitions, we can model who is going to achieve a property (a coalition can be an individual entity or even an empty set -modeling ‘nature’-). An example about social choice (voting) mechanism. Now, mechanisms can be validated.The logic can capture dependencies among agents, as stressfulness (all goals met), veto (j needs i to achieve its goal), mutual dependence (all agents are mutual dependence… veto relationship)

(note: but we can’t model actions yet, so I guess it isn’t useful for us)

A concrete application about social laws (normative systems). Objectives will be ATL formulae \(phi\) and mechanisms are behavioral constraints \(\beta\) To avoid undesirable behaviors, we have to cut out some transitions. An effective social law \((\phi,\beta) \models \phi\). But compute this is a NP-hard problem. An example with the typical train organization in a tunnel. But you cannot model just the properties you want to avoid. The properties you want to preserve have to be modeled too in order to have system doing useful things.

But, what to do with non-compliance? The idea isto incentive compatibility and, to do this, we need preferences (a prioritized list of goal formulae). I like this idea: the utility of the agent comes from this list, from a worst (and weak) rule to the best (and stronger) rule. For instance, related with resources, have it assigned often and for a long time.

Blogged with the Flock Browser

[AT Workshop] Session 1

Towards the biological basis of cooperation
(Arcadi Navarro)

Talking about genome and human evolution. The interesting thing: the effects on social behavior.

After a very interesting introduction to genomic, begins trying to relate genetics with social behavior: because to cooperate can have some explanations in our genes (and this can be the explanation of why humans have  been a successful specie): genetic variability for behavioral traits is considerable. The problem is that this is very difficult to interpret. Fortunately, there are some genetics related with economic behavior that can be studied and replicated in labs.

Example: the ultimate game: people trend to make 50:50 offers and to reject less that 30% (not an reasonable decision from an economic point of view). But chimpanzees behaves as rational maximizes in an ultimatum game. Both species have evolve completely different behaviors. Why? we have to study this from a genetic perspective. -> agents playing games are as chimpanzees. And researchers are discovered that serotonin makes individuals to be more generous (just a joke: men have more serotonin than women). Or even between MZ twins, differences in the acceptance threshold in ultimatum game have been observed. Examples with more genes.

Measuring Strategic Uncertainly and Risk in Coordination-, entry-Games and lotteries with fMRI
(Rosemari Nagel)

Uncertainty can be classified as

  • exogenous (risk): know the prob. of all possible states of the world (objective prob.)
  • endogenous: in absence of endogenously given prob.;  -> strategic uncertainty (SU) e.g. outcomes depends on social interaction -games- (subjective prob.)

How brain solve individual or strategic uncertainty? Can we predict choices and brain activity in games?
Results: people behaves similarly in lottery and coordination games, but not in entry games. And the activity in the brain increases in lottery -> coordination -> entry. Some graphics about the different parts of the brain active while playing each type of game. Similar activity in entry games of risk lovers and risk averse people.

Summarizing, the entry games create mode strategic uncertainty as predicted by the nature of the mixed equilibrium which also involves levels of reasoning.

Blogged with the Flock Browser

mWater: a case study of AT

mWater es el demostrador que tenemos asignado en el AT. Se trata de ver cómo usar los conceptos de acuerdos en un caso de gestión eficiente de agua.

Se aplica sobre un entorno en concreto (acuífero La Mancha Oriental), donde se identifica además la estructura de distribución del agua y la estructura social existente. La implementación se realiza sobre instituciones electrónicas (primer prototipo) y Thomas (sistemas abiertos). Actualmente se está desarrollando en mercado intra-cuenca para la gestión de derechos de agua modelada con Islander.

Básicamente, el proceso puede resumirse de la siguiente manera

  1. cuando los agentes entran en el mercado, deben registrarse en las escenas iniciales
  2. a continuación, pasan a un trading hall en el que pueden ver qué transacciones se están realizando
  3. las transacciones se realizan en un conjunto de trading tables. El agente selecciona en qué mesa(s) quiere participar.
  4. en las mesas se firman contratos entre partes donde se negocian derechos de agua
  5. los contratos deben quedar expuestos 30 días para que otros agentes afectados puedan efecturar quejas y alegaciones (grievance)
  6. Una vez firmado el contrato, cualquier incidencia se resuelve mediante un mecanismo de quejas.

mWater es un entorno regulado y esta regulación se realiza a través de normas. Son de 3 tipos

  • definidas por el gobierno (p.ej. PHN)
  • definidas por las asociaciones de regantes
  • normas sociales

Otro aspecto importante es la gestión organizativa de los agetes que participan en cada escenario: cómo comparten normas, como se participa en actividades reguladas, emergencia y consolidación de normas comunes, toma de decisiones colaborativa y estudio de la dinámica de la propia organización.

Dado que la gestión de conflictos se realiza mediante quejas, es neceario definir las tecnologías necesarias: protocolos, artuitecturas, ténicas de negociación y argumentación… Los conflictos surgen del hecho de que algún participante no cumple su parte del contrato, por lo que habrá que poder validarlos y detectar estas situaciones. Todo esto se instrumenta a través de acuerdos.

Acaba explicando qué tareas del AT están relacionadas con qué partes del mWater. Un poco liado para contarlo aquí… pero podéis verlo en las transparenicas.

Blogged with the Flock Browser