<?xml version="1.0" standalone="yes"?>
<Paper uid="W98-1212">
  <Title>I l I I | l / A Bayesian Approach to Automating Argumentation</Title>
  <Section position="4" start_page="0" end_page="0" type="intro">
    <SectionTitle>
USER
</SectionTitle>
    <Paragraph position="0"> work structure with nodes that represent propositions, and connecting links that represent the inferences that connect these propositions. An Argument Graph is fleshed out by consulting several sources of information called Reasoning Agents and incorporating the relevant inferences and propositions returned by these sources into the Argument Graph. This Argument Graph is passed to the Strategist.</Paragraph>
    <Paragraph position="1"> The Strategist decides what NAG should do next: call the Generator to continue the argument building process; call the Analyzer (Section 4) to estimate how nice the current Argument Graph is; or present an argument based on the current Argument Graph to the user for inspection and response.</Paragraph>
    <Paragraph position="2"> The Strategist will pass the current Argument Graph to the Analyzer at least once before the argument is presented to the user, and often more than once. To estimate the persuasive power of an argument represented by an Argument Graph, the Analyzer consults a revisable user model that reflects the beliefs and cognitive abilities of the audience.</Paragraph>
    <Paragraph position="3"> The Analyzer uses a normative model to gauge the normative strength of an argument. Belief updating in both the user and the normative model is done by a constrained Bayesian propagation scheme (Sec-McConachy, Korb and Zukerman 91 A Bayesian Approach to Automating Argumentation Richard McConachy, Kevin B. Korb and Ingrid Zukerman (1998) A Bayesian Approach to Automating Argumentation. In D.M.W. Powers (ed.) NeMLaP3/CoNLL98: New Methods in Language Processing and Computational Natural Language Learning, ACL, pp 91-100.</Paragraph>
    <Paragraph position="4"> tion 4). In the user model, Bayesian updating is adjusted by multiplicative factors which model three human cognitive weaknesses (Section 4.1).</Paragraph>
    <Paragraph position="5"> If the Analyzer detects problems with the Argument Graph it highlights the weaknesses for the Generator to fix. In this way a cycle of alternately critiquing and extending the graph is continued until a successful Argument Graph is built, or NAG is confronted with an event that prevents it from continuing, such as the Generator failing to find relevant new evidence or the Strategist noticing that the allowed amount of time has run out.</Paragraph>
  </Section>
class="xml-element"></Paper>