Subtopic Deep Dive

Argumentation Frameworks in AI
Research Guide

What is Argumentation Frameworks in AI?

Argumentation frameworks in AI are formal structures for representing and evaluating arguments in multi-agent systems, enabling defeasible reasoning and negotiation under incomplete knowledge.

Dung's abstract frameworks form the basis, extended by structured and value-based semantics (Prakken, 2010; Bench-Capon, 2003). Key formalisms include Defeasible Logic Programming (DeLP) by García and Simari (2004, 814 citations) and argumentation-theoretic default reasoning by Bondarenko et al. (1997, 665 citations). Over 10 highly cited papers from 1997-2010 define the field with 500+ citations each.

15
Curated Papers
3
Key Challenges

Why It Matters

Argumentation frameworks enable agents to negotiate by exchanging arguments, resolving conflicts in multi-agent systems (Parsons, 1998; Rahwan et al., 2003). Value-based frameworks support persuasion in practical reasoning for decision support (Bench-Capon, 2003). DeLP integrates logic programming with defeasible rules for legal and ethical agent deliberation (García and Simari, 2004). Applications include autonomous agreement reaching (Kraus et al., 1998).

Key Research Challenges

Scalability in Complex Graphs

Evaluating acceptability in large argumentation graphs with thousands of nodes remains computationally expensive. Dung frameworks lack efficient semantics for real-time multi-agent negotiation (Prakken, 2010). Bondarenko et al. (1997) highlight default reasoning bottlenecks.

Preference Handling

Incorporating agent preferences into semantics leads to multiple stable extensions without clear selection. Value-based approaches address this partially but struggle with dynamic preferences (Bench-Capon, 2003). Rahwan et al. (2003) note negotiation protocol gaps.

Structured Argument Integration

Mapping natural language arguments to structured frameworks loses expressive power. Prakken (2010) proposes abstractions but verification against real disputes is challenging. Parsons (1998) identifies implementation hurdles in agent systems.

Essential Papers

1.

Defeasible logic programming: an argumentative approach

Alejandro Javier García, Guillermo Ricardo Simari · 2004 · Theory and Practice of Logic Programming · 814 citations

The work reported here introduces Defeasible Logic Programming (DeLP), a formalism that combines results of Logic Programming and Defeasible Argumentation. DeLP provides the possibility of represen...

2.

Persuasion in Practical Argument Using Value-based Argumentation Frameworks

Trevor Bench‐Capon · 2003 · Journal of Logic and Computation · 751 citations

In many cases of disagreement, particularly in situations involving practical reasoning, it is impossible to demonstrate conclusively that either party is wrong. The role of argument in such cases ...

3.

An abstract, argumentation-theoretic approach to default reasoning

A. G. Bondarenko, Phan Minh Dũng, Robert Kowalski et al. · 1997 · Artificial Intelligence · 665 citations

4.

An abstract framework for argumentation with structured arguments

Henry Prakken · 2010 · Argument & Computation · 661 citations

An abstract framework for structured arguments is presented, which instantiates Dung's (‘On the Acceptability of Arguments and its Fundamental Role in Nonmonotonic Reasoning, Logic Programming, and...

5.

Agents that reason and negotiate by arguing

Simon Parsons · 1998 · Journal of Logic and Computation · 657 citations

The need for negotiation in multi-agent systems stems from the requirement for agents to solve the problems posed by their interdependence upon one another. Negotiation provides a solution to these...

6.

Argumentation-based negotiation

Iyad Rahwan, Sarvapali D. Ramchurn, Nicholas R. Jennings et al. · 2003 · The Knowledge Engineering Review · 596 citations

Negotiation is essential in settings where autonomous agents have conflicting interests and a desire to cooperate. For this reason, mechanisms in which agents exchange potential agreements accordin...

7.

Logical models of argument

Carlos Iván Chesñevar, Ana Gabriela Maguitman, Ronald P. Loui · 2000 · ACM Computing Surveys · 520 citations

Logical models of arguement formalize commonsense reasoning while taking process and computation seriously. This survey discusses the main ideas that characterize different logical models of argume...

Reading Guide

Foundational Papers

Start with Prakken (2010) for structured Dung instantiation and García and Simari (2004) DeLP for logic programming integration; then Bench-Capon (2003) values and Parsons (1998) agent negotiation.

Recent Advances

Prakken (2010) structured frameworks and Rahwan et al. (2003) negotiation protocols represent key pre-2015 advances with 661 and 596 citations.

Core Methods

Dung abstract graphs with acceptability semantics; DeLP defeasible rules; value-based ordering; argumentation protocols for multi-agent exchange (Bondarenko et al., 1997; Kraus et al., 1998).

How PapersFlow Helps You Research Argumentation Frameworks in AI

Discover & Search

Research Agent uses citationGraph on García and Simari (2004) to map DeLP citations, revealing 814 connected works; exaSearch queries 'Dung frameworks multi-agent negotiation' for 250M+ OpenAlex papers; findSimilarPapers extends to Rahwan et al. (2003).

Analyze & Verify

Analysis Agent applies readPaperContent to Prakken (2010) for structured argument details, verifyResponse (CoVe) checks semantics claims against Bondarenko et al. (1997); runPythonAnalysis simulates Dung graph extensions with NetworkX, GRADE scores defeasible inference evidence.

Synthesize & Write

Synthesis Agent detects gaps in preference semantics across Bench-Capon (2003) and Rahwan et al. (2003); Writing Agent uses latexEditText for framework diagrams, latexSyncCitations integrates 10 key papers, latexCompile generates negotiation protocol reports, exportMermaid visualizes attack graphs.

Use Cases

"Compare DeLP with value-based argumentation in agent negotiation"

Research Agent → searchPapers('DeLP negotiation') → Analysis Agent → readPaperContent(García 2004, Bench-Capon 2003) → runPythonAnalysis(acceptability comparison) → GRADE table of extensions.

"Generate LaTeX figure of Dung argumentation graph from Parsons 1998"

Research Agent → citationGraph(Parsons 1998) → Synthesis → gap detection → Writing Agent → latexGenerateFigure(Dung graph) → latexSyncCitations(5 papers) → latexCompile(PDF with attack/defeat edges).

"Find GitHub code for argumentation-based multi-agent negotiation"

Research Agent → findSimilarPapers(Rahwan 2003) → Code Discovery → paperExtractUrls → paperFindGithubRepo → githubRepoInspect → runPythonAnalysis(test negotiation protocols).

Automated Workflows

Deep Research scans 50+ papers from García-Simari (2004) citations for systematic DeLP review: searchPapers → citationGraph → DeepScan 7-steps → structured report. Theorizer generates value-based negotiation theory: readPaperContent(Bench-Capon 2003) → contradiction flagging → theory synthesis. DeepScan verifies Kraus et al. (1998) implementation with CoVe checkpoints on agreement models.

Frequently Asked Questions

What defines argumentation frameworks?

Abstract structures modeling arguments as nodes with attack relations, per Dung semantics instantiated by Prakken (2010). Extensions include stable, preferred, grounded semantics.

What are main methods?

DeLP combines logic programming with defeasible rules (García and Simari, 2004). Value-based frameworks rank arguments by audience values (Bench-Capon, 2003). Negotiation protocols exchange arguments (Rahwan et al., 2003).

What are key papers?

García and Simari (2004, 814 citations) on DeLP; Bench-Capon (2003, 751 citations) on value-based persuasion; Prakken (2010, 661 citations) on structured arguments; Parsons (1998, 657 citations) on arguing agents.

What open problems exist?

Efficient semantics for dynamic preferences; scaling to real-time multi-agent systems; bridging structured arguments with natural language (Prakken, 2010; Rahwan et al., 2003).

Research Multi-Agent Systems and Negotiation with AI

PapersFlow provides specialized AI tools for Computer Science researchers. Here are the most relevant for this topic:

See how researchers in Computer Science & AI use PapersFlow

Field-specific workflows, example queries, and use cases.

Computer Science & AI Guide

Start Researching Argumentation Frameworks in AI with AI

Search 474M+ papers, run AI-powered literature reviews, and write with integrated citations — all in one workspace.

See how PapersFlow works for Computer Science researchers