<?xml version="1.0" standalone="yes"?>
<Paper uid="N04-4024">
  <Title>Direct Maximization of Average Precision by Hill-Climbing, with a Comparison to a Maximum Entropy Approach</Title>
  <Section position="6" start_page="0" end_page="0" type="evalu">
    <SectionTitle>
5 Results
</SectionTitle>
    <Paragraph position="0"> The average precision scores obtained by the maximum entropy and weight search algorithm experiments are listed in Table 1. The Best AP and No. Terms columns describe the query size at which average precision was best and the score at that point. These columns show that the maximum entropy approach performs just as well as the average precision hill-climber, and in some cases actually performs slightly better. This suggests that the metric divergence as seen in Figure 1 did not prohibit the maximum entropy approach from maximizing average precision in the course of maximizing likelihood.</Paragraph>
    <Paragraph position="1"> The 5 term AP column compares the performance of the algorithms on smaller queries. The weight search algorithm shows a slight advantage over the maximum entropy model on 10 of the 15 topics and equal performance on the others, but de nitive conclusions are dif cult at this stage.</Paragraph>
    <Paragraph position="2"> Figure 3 shows the average precision achieved by the weight search algorithm, for all 20 query sizes and for all 15 topics. Unlike the maximum entropy results, the algorithm is guaranteed to yield monotonically non-decreasing scores.</Paragraph>
    <Paragraph position="3"> Topic 5 term AP Best AP No. Terms</Paragraph>
  </Section>
class="xml-element"></Paper>